Search

In aquaculture, accurately counting fish fries is a prerequisite for fish population management and marketing. However, due to the fry overlapping and occlusion issues, the manual counting method is time-consuming, and the counting result might be inaccurate. Therefore, we utilized the computer vision technique to develop a lightweight fish fry counting model to address this issue. First, we constructed a diverse dataset containing images of largemouth bass in varying numbers, captured under different lighting conditions and water depths. Then, we optimized the YOLOv5s model by channel pruning to reduce the model size, while maintaining the detection accuracy. Through extensive experiments, we examined the effect of different pruning rates on the model performance and compared the pruned YOLOv5s model with state-of-the-art detection models through the evaluation criteria like precision, recall, mean average precision (mAP), model size, GFLOPs, and detection speed. In addition, we investigated the impact of environmental factors, such as lighting conditions and water depths, on the detection performance of the pruned YOLOv5s model. The experimental results demonstrated that the YOLOv5s model with a pruning rate of 15% achieved over 90% accuracy and 13 FPS in the dense and complex scenes, which met the practical requirement for the fry counting task. In addition, we also identified that the pruned YOLOv5s model would achieve the optimal performance under the white illumination and shallow water depth setting. In conclusion, this study provided an efficient solution for fish fry counting, and the proposed model was expected to be applied in the real-world application. [ABSTRACT FROM AUTHOR]

In monitoring transmission line external damage prevention, due to the limited memory computing power of the equipment, the image needs to be transmitted to the data center at regular intervals, resulting in a high false negative rate. Therefore, this paper proposes a target detection method based on lightweight YOLOv5s. First, DSConv and improved E‐ELAN are used in Backbone to reduce the model's parameters. Then, GSConv and VoV‐GSCSP are introduced in Neck to reduce the complexity of the model. Finally, the Mish activation function achieves more effective feature transfer. According to the experimental findings, the proposed model's parameters are about 37% smaller than the original model's, and the calculation amount is about 53% smaller. The detection accuracy on the self‐built data set is the same, which proves that the proposed algorithm can reduce the model while maintaining high detection performance. It has specific practical significance for the terminal real‐time detection of external mechanical damage targets. © 2024 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

Aiming at the current coal gangue recognition model’s difficulty in recognizing aggregated coal gangue targets, which leads to the problems of misdetections, omissions, and inaccurate localization, we propose a coal gangue recognition method based on the SFD-YOLOv5s model. The SE attention mechanism is improved in incorporating maximum pooling and introduced into the backbone network of the YOLOv5s model to improve the model’s ability to extract the critical feature information of the coal gangue. Furthermore, to improve the recognition ability of the YOLOv5s model for aggregated targets, we introduce the Focal-EIoU loss function and the Dyhead. The experimental results show that the overall performance of the SFD-YOLOv5s model has surpassed other mainstream object detection models of the YOLO series. The SFD-YOLOv5s model has the highest precision, recall, and mean average precision. The precision reaches 97.9%, the recall is 96.2%, and the mean average precision is 94.5%, which is 8.7%, 8.9%, and 9.0% higher compared with the YOLOv5s model, respectively. The FPS is 96.15, which can fully meet the demand for real-time detection of coal gangue. Meanwhile, the SFD-YOLOv5s model has the best recognition effect of aggregated coal gangue targets. [ABSTRACT FROM AUTHOR]

Nursery planting is ever expanding at present, particularly with the increasing demand for fruit and ornamental trees. Simple tools cannot fully meet the efficient work of large-scale production, due to the labor intensity and pesticide demand. Alternatively, spray robots can be expected to reduce the large number of tasks in modern agriculture. However, it is still lacking in the recognition accuracy and target identification of spray robots in nurseries. In this study, a nursery target detection was proposed using an improved YOLOv5s. Firstly, partial convolution (PConv) was introduced into the comprehensive convolution block (C3) to reduce the computational complexity, in order to improve the backbone network. A coordinate attention mechanism was added at the highest dimensional feature to enhance location awareness. Secondly, the neck structure of the network model was optimized to enhance the feature extraction of the model. At the same time, the bilinear interpolation was used for up-sampling operations during training. Finally, the original coupling detection head was replaced with the improved light decouple one, in order to further improve the detection accuracy. A nursery dataset was constructed with a total of 2 000 pictures, including three representative objects: trees, pedestrians, and cultivation pots. The images of the objects were collected from pedestrians with various postures at varying distances (far, medium, and near). The dataset included a total of 5 769 trees, 1688 pedestrians, and 6 178 pots. Only the trees and pots were marked in the first row during labeling, according to the operation requirements. Once the pedestrians are detected, they must be identified to consider the safety of pedestrians. Therefore, all pedestrians were labeled regardless of distance. The experimental results show that the detection of the model was improved differently via the C3 module and coordinate attention mechanism, as well as the optimized neck network structure and up-sampling mode using a lightweight decoupling head. Finally, the average mAP0.5, mAP0.5:0.95, accuracy, and recall reached 88.2%, 54.7%, 86.0%, and 82.4%, respectively. The size of the improved network model was 14.1 MB, the average detection speed of a single image was 19.5 ms, and the average frame rate was 51.3 frames. The mAP0.5, mAP0.5:0.95, accuracy, and recall increased by 4.6, 5.9, 1.8 and 3.4 percentage points, respectively, compared with the original YOLOv5s. The improved network had the highest accuracy, mAP0.5 and mAP0.5:0.95, compared with the current mainstream single-stage target detection YOLOv3-tiny, YOLOv7-tiny, and the latest YOLOv8s model. The experimental results show that the improved model performed better in accurately and rapidly identifying the pedestrians, pots, and trees in a complex environment. The research findings can also provide technical support to the operational activities of electric spraying robots in nurseries. [ABSTRACT FROM AUTHOR]

Abnormal behavior of crew members is an important cause of frequent ship safety accidents. The existing abnormal crew recognition algorithms are affected by complex ship environments and have low performance in real and open shipborne environments. This paper proposes an abnormal crew detection network for complex ship scenarios (ACD-Net), which uses a two-stage algorithm to detect and identify abnormal crew members in real-time. An improved YOLOv5s model based on a transformer and CBAM mechanism (YOLO-TRCA) is proposed with a C3-TransformerBlock module to enhance the feature extraction ability of crew members in complex scenes. The CBAM attention mechanism is introduced to reduce the interference of background features and improve the accuracy of real-time detection of crew abnormal behavior. The crew identification algorithm (CFA) tracks and detects abnormal crew members' faces in real-time in an open environment (CenterFace), continuously conducts face quality assessment (Filter), and selects high-quality facial images for identity recognition (ArcFace). The CFA effectively reduces system computational overhead and improves the success rate of identity recognition. Experimental results indicate that ACD-Net achieves 92.3% accuracy in detecting abnormal behavior and a 69.6% matching rate for identity recognition, with a processing time of under 39.5 ms per frame at a 1080P resolution. [ABSTRACT FROM AUTHOR]

Mildew in maize kernels is typically caused by various fungi, necessitating prompt detection and treatment to minimize losses during harvest and storage. In this study, a deep learning YOLOv5s algorithm based on machine vision technology was employed to develop a maize seed surface mildew detection model and to enhance its portability for deployment on additional mobile devices. To guarantee the fruitful progression of this research, an initial experiment was conducted on maize seeds to obtain a sufficient number of images of mildewed maize kernels, which were classified into three grades (sound, mild, and severe). Subsequently, a maize seed image was extracted to create an image of a single maize seed, which was then divided to establish the data set. An enhanced YOLOv5s–ShuffleNet–CBAM model was ultimately developed. The results demonstrated that the model achieved with an mAP50 value of 0.955 and a model size of 2.4 MB. This resulted in a notable reduction in the model parameters and calculation amount while simultaneously enhancing model precision. Furthermore, K-fold cross-validation demonstrated the model stability, and Grad-CAM validated the model effectiveness. In the future, the proposed lightweight model in this study can be applied to other crops in the context of portable or online inspection systems, thus advancing effective and high-quality agricultural applications. [ABSTRACT FROM AUTHOR]

In order to improve the target visual recognition and localization accuracy of robotic arms in complex scenes with similar targets, hybrid recognition and localization methods based on an industrial camera and depth camera are proposed. First, according to the speed and accuracy requirements of target recognition and localization, YOLOv5s is introduced as the basic algorithm model for target hybrid recognition and localization. Then, in order to improve the accuracy of target recognition and coarse localization based on an industrial camera (eye-to-hand), the AFPN feature fusion module, simple and parameter-free attention module (SimAM), and soft non-maximum suppression (Soft NMS) are introduced. In order to improve the accuracy of target recognition and fine localization based on a depth camera (eye-in-hand), the SENetV2 backbone network structure, dynamic head module, deformable attention mechanism, and chain-of-thought prompted adaptive enhancer network are introduced. After that, on the basis of constructing a dual camera platform for target hybrid recognition and localization, the hand–eye calibration, collection and production of image datasets required for model training are completed. Finally, for the docking of the oil filling port, the hybrid recognition and localization experimental tests are completed in sequence. The test results show that in target recognition and coarse localization based on the industrial camera, the recognition accuracy of the designed model reaches 99%, and the average localization errors in the horizontal and vertical directions are 2.22 mm and 3.66 mm, respectively. In target recognition and fine localization based on the depth camera, the recognition accuracy of the designed model reaches 98%, and the average errors in depth, horizontal, and vertical directions are 0.12 mm, 0.28 mm, and 0.16 mm, respectively. These not only verify the effectiveness of the target hybrid recognition and localization methods based on dual cameras, but also demonstrate that they meet the high-precision recognition and localization requirements in complex scenes. [ABSTRACT FROM AUTHOR]

Identifying and statistically analyzing soybean pod types are crucial for seed evaluation and yield estimation. Traditional visual assessment by breeding personnel is time-consuming, labor-intensive, and prone to subjective bias, especially with large datasets. Automatic assessment methods usually struggle with the highly confusing pod types with two and three seeds, affecting the model's identification accuracy. To address these issues, we propose to improve the standard YOLOv5s object detection model to enhance the differentiation between pod types and to boost the model's efficiency in prediction. To reduce the number of parameters and the computational load, we propose to introduce the FasterNet Block module in the FasterNet model into the original C3 module, leading to improvements in both detection accuracy and speed. To strengthen the feature extraction and representation for specific targets, the Efficient Multi-Scale Attention (EMA) module is incorporated into the C3 module of the backbone network, improving the identification of similar pod types. The Inner-IoU is combined with the CIoU as the loss function to further enhance detection accuracy and generalization. Experiments comparing FEI-YOLO with the baseline YOLOv5s show that FEI-YOLO achieves an m A P @ 0.5 of 98.6% and an m A P @ 0.5 : 0.95 of 81.1%, with improvements of 1.5% and 1.4%, respectively. Meanwhile, the number of parameters is reduced by 13.2%, and FLOPs decreased by 10.8%, demonstrating the model's effectiveness and efficiency, enabling rapid and accurate identification of soybean pod types from images. [ABSTRACT FROM AUTHOR]

To address the challenges of low detection accuracy, slow detection speed, coarse feature extraction, and the difficulty of detection deployment in complex forest fire backgrounds, this paper presents a forest fire object detection algorithm based on an improved YOLOv5s (YOLOv5s-ACE). The algorithm not only realizes the accurate identification of small objects, but also guarantees the accuracy and speed of detection. Firstly, YOLOv5s-ACE uses Copy-Pasting data enhancement to expand the small object sample set to reduce the overfitting risk in the process of model training. Secondly, it choose Atrous Spatial Pyramid Pooling (ASPP) to replace Spatial Pyramid Pooling (SPP) module in backbone part of YOLOv5 network. Therefore, the proposed algorithm can enlarge the receptive field while ensuring the resolution, which is conducive to the accurate positioning of small object forest flame. Third, after adding the Convolutional Block Attention Module (CBAM) module to the C3 module of the Neck layer, the key features of the forest flame object can be further screened, while irrelevant information that interferes with the flame detection, such as background information, can be eliminated. The network performance of forest fire detection is improved without increasing the depth, width and resolution of the input image. Finally, we replace CIOU losses (Complete-IoU) with EIOU losses (Efficient-IoU) to optimize the performance of the model and improve accuracy. The experimental results show that compared with the original algorithm, the proposed object detection algorithm improves mean Average Precision (mAP) by 5.6%, Precision by 2.7%, Recall by 6.5% and GFlops by 6.7%. Even compared with the YOLOv7 algorithm, the proposed algorithm YOLOv5s-ACE increases mAP by 0.9%, Precision by 2.2%, and Recall by 0.3%. [ABSTRACT FROM AUTHOR]

This paper presents significant improvements in the accuracy and computational efficiency of safety helmet detection within industrial environments through the optimization of the you only look once version 5 small (YOLOv5s) model structure and the enhancement of its loss function. We introduce the convolutional block attention module (CBAM) to bolster the model's sensitivity to key features, thereby enhancing detection accuracy. To address potential performance degradation issues associated with the complete intersection over union (CIoU) loss function in the original model, we implement the modified penalty-decay intersection over union (MPDIoU) loss function to achieve more stable and precise bounding box regression. Furthermore, considering the original YOLOv5s model's large parameter count, we adopt a lightweight design using the MobileNetV3 architecture and replace the original squeeze-and-excitation (SE) attention mechanism with CBAM, significantly reducing computational complexity. These improvements reduce the model's parameters from 15.7 GFLOPs to 5.7 GFLOPs while increasing the mean average precision (mAP) from 82.34% to 91.56%, demonstrating its superior performance and potential value in practical industrial applications. [ABSTRACT FROM AUTHOR]

Simple Summary: The aim of this study was to monitor the rumination behavior of dairy cows by combining the YOLOv5s object detection and DeepSort tracking algorithms. To verify the applicability of the algorithmic model in real dairy farm settings, ruminating cows exhibiting various behavioral disturbances were tested, and the model was comprehensively evaluated using multiple performance metrics. Additionally, two methods for determining the chewing actions of dairy cows were proposed, and a comparative analysis of the false detection rates for the chewing counts was conducted. The results demonstrate that the algorithmic system was effective in detecting and tracking the upper and lower jaws of the cows, with a low false detection rate for the chewing counts. To address behavioral interferences such as head turning and lowering during rumination in group-housed dairy cows, an enhanced network algorithm combining the YOLOv5s and DeepSort algorithms was developed. Initially, improvements were made to the YOLOv5s algorithm by incorporating the C3_CA module into the backbone to enhance the feature interaction and representation at different levels. The Slim_Neck paradigm was employed to strengthen the feature extraction and fusion, and the CIoU loss function was replaced with the WIoU loss function to improve the model's robustness and generalization, establishing it as a detector of the upper and lower jaws of dairy cows. Subsequently, the DeepSort tracking algorithm was utilized to track the upper and lower jaws and plot their movement trajectories. By calculating the difference between the centroid coordinates of the tracking boxes for the upper and lower jaws during rumination, the rumination curve was obtained. Finally, the number of rumination chews and the false detection rate were calculated. The system successfully monitored the frequency of the cows' chewing actions during rumination. The experimental results indicate that the enhanced network model achieved a mean average precision (mAP@0.5) of 97.5% and 97.9% for the upper and lower jaws, respectively, with precision (P) of 95.4% and 97.4% and recall (R) of 97.6% and 98.4%, respectively. Two methods for determining chewing were proposed, which showed false detection rates of 8.34% and 3.08% after the experimental validation. The research findings validate the feasibility of the jaw movement tracking method, providing a reference for the real-time monitoring of the rumination behavior of dairy cows in group housing environments. [ABSTRACT FROM AUTHOR]

In order to improve the performance of potato planter, reduce miss-seeding rates, enhance the overall quality of the seeding operation, and ultimately increase the yield of the potato, it is necessary to implement effective technical means to monitor and identify the miss-seeding issues during the seeding process. The existing miss-seeding detection technologies commonly use sensors to monitor, but such technologies are easily affected by factors like heavy dust and strong vibrations, resulting in poor interference resistance and adaptability. Therefore, this study aims to explore and apply deep learning algorithms to achieve real-time monitoring of the miss-seeding phenomenon in potato planter during the planting process. Considering both the lightweight of the miss-seeding detection model and its practical deployment, this study selects and adapts the YOLOv5s algorithm to achieve this goal. Firstly, the attention mechanism is integrated into the backbone network to suppress background interference and improve detection accuracy. Secondly, the non-maximum suppression algorithm is improved by replacing the original IoU-NMS with the Soft-NMS algorithm to enhance the bounding box regression rate and reduce missed detections of potato seeds due to background overlap or occlusion. Experimental results show that the accuracy of the improved algorithm in detecting miss-seeding increased from 96.02% to 98.30%, the recall rate increased from 96.31% to 99.40%, and the mean average precision (mAP) improved from 99.12% to 99.40%. The improved model reduces missed and false detections, provides more precise target localization, and is suitable for miss-seeding detection in natural environments for potato planter, providing technical and theoretical support for subsequent intelligent reseeding in potato planter. [ABSTRACT FROM AUTHOR]

Fusarium head blight (FHB) is one of the most destructive diseases in global wheat production. In order to count the FHB-infected wheat ears under field conditions, this study proposed an algorithm for diseased wheat ear detection based on improved YOLOv5s (Tr-YOLOv5s). The Swin Transformer was used to replace the CSPDarknet backbone network to enhance the extraction of characteristic information of the population wheat ears of FHB in the field background. The convolutional block attention module (CBAM) attention mechanism was added to improve the detection effect of target wheat ears, subsequently improving the overall accuracy of the model. The original loss function complete intersection over union (CIoU) was replaced by Scylla intersection over union (SIoU) loss to accelerate the model convergence and decrease the loss value. The results showed that the mean average precision (mAP) of the Tr-YOLOv5s model reached 90.64%, making a 4.63% improvement compared to the original YOLOv5s model. The improved model could quickly detect and count wheat FHB ear in the field environment, which laid a foundation for the subsequent automatic disease identification and grading of wheat FHB under field conditions. [ABSTRACT FROM AUTHOR]

Sugarcane smut, a serious disease caused by the fungus Sporosorium scitamineum, can result in 30% to 100% cane loss. The most affordable and efficient measure of preventing and handling sugarcane smut disease is to select disease-resistant varieties. A comprehensive evaluation of disease resistance based on the incidence of smut disease is essential during the selection process, necessitating the rapid and accurate identification of sugarcane smut. Traditional identification methods, which rely on visual observation of symptoms, are time-consuming, costly, and inefficient. To address these limitations, we present the lightweight sugarcane smut detection model (YOLOv5s-ECCW), which incorporates several innovative features. Specifically, the EfficientNetV2 is incorporated into the YOLOv5 network to achieve model compression while maintaining high detection accuracy. The convolutional block attention mechanism (CBAM) is added to the backbone network to improve its feature extraction capability and suppress irrelevant information. The C3STR module is used to replace the C3 module, enhancing the ability to capture global large targets. The WIoU loss function is used in place of the CIoU one to improve the bounding box regression's accuracy. The experimental results demonstrate that the YOLOv5s-ECCW model achieves a mean average precision (mAP) of 97.8% with only 4.9 G FLOPs and 3.25 M parameters. Compared with the original YOLOv5, our improvements include a 0.2% increase in mAP, a 54% reduction in parameters, and a 70.3% decrease in computational requirements. The proposed model outperforms YOLOv4, SSD, YOLOv5, and YOLOv8 in terms of accuracy, efficiency, and model size. The YOLOv5s-ECCW model meets the urgent need for the accurate real-time identification of sugarcane smut, supporting better disease management and selection of resistant varieties. [ABSTRACT FROM AUTHOR]

To address the issue of the traditional fish hypoxia detection methods, which have low accuracy and require a lot of labor, a Prune-YOLOv5s-based hypoxia risk assessment method for farmed fish is proposed. This paper introduces a hypoxia risk assessment method for cultured fish based on the Prune - YOLOv5s algorithm. This method first collects data on aquatic surface respiration (ASR) performed by fish under hypoxic conditions to create a data set for fish hypoxia. The dataset is then utilized to train the YOLOv5s model. Then, the lightweight and improved YOLOv5s model was used to monitor the behavior of fish surface respiration during hypoxia in real-time. The introduction of the ASR coefficient allows for quantifying ASR instances in fish, which indicates hypoxia risk. The fish hypoxia assessment module is designed to evaluate the risk of hypoxia. The improved performance of the YOLOv5s model before and after modifications and the accuracy of the fish hypoxia assessment module are tested through the fish hypoxia experiment. The test results show that compared with the YOLOv5s model, the detection accuracy, model size, inference speed, and detection speed of the PruneYOLOv5s model have been significantly improved. Among them, the detection accuracy of the 65% Prune YOLOv5s model, which has the best comprehensive performance, has been increased by 0.6% compared with the original model. The size of the model is reduced to 45. 3% of the original model. The inference speed is improved by 23.8%, and the detection speed is enhanced by 31.4%. The fish hypoxia assessment method achieves 97.4% accuracy in the test set of 39 test videos and performs well in the hypoxia cycle experiment. The research indicates that the Prune-YOLOv5s-based hypoxia risk assessment method for cultured fish can effectively detect hypoxic conditions and provide accurate risk alerts, showing high feasibility for practical application. [ABSTRACT FROM AUTHOR]

Copyright of Packaging & Food Machinery is the property of Packaging & Food Machinery Magazine and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This work was conducted mainly to provide a healthy and safe monitoring system for the elderly living in the home environment. In this paper, two different target fall detection schemes are proposed based on whether the target is visible or not. When the target is visible, a vision-based fall detection algorithm is proposed, where an image of the target captured by a camera is transmitted to the improved You Only Look Once version 5s (YOLOv5s) model for posture detection. In contrast, when the target is invisible, a WiFi-based fall detection algorithm is proposed, where channel state information (CSI) signals are used to estimate the target's posture with an improved long short-term memory (LSTM) model. In the improved YOLOv5s model, adaptive picture scaling technology named Letterbox is used to maintain consistency in the aspect ratio of images in the dataset, and the weighted bidirectional feature pyramid (BiFPN) and the attention mechanisms of squeeze-and-excitation (SE) and coordinate attention (CA) modules are added to the Backbone network and Neck network, respectively. In the improved LSTM model, the Hampel filter is used to eliminate the noise from CSI signals and the convolutional neural network (CNN) model is combined with the LSTM to process the image made from CSI signals, and thus the object of the improved LSTM model at a point in time is the analysis of the amplitude of 90 CSI signals. The final monitoring result of the health status of the target is the result of combining the fall detection of the improved YOLOv5s and LSTM models with the physiological information of the target. Experimental results show the following: (1) the detection precision, recall rate, and average precision of the improved YOLOv5s model are increased by 7.2%, 9%, and 7.6%, respectively, compared with the original model, and there is almost no missed detection of the target; (2) the detection accuracy of the improved LSTM model is improved by 15.61%, 29.36%, and 52.39% compared with the original LSTM, CNN, and neural network (NN) models, respectively, while the convergence speed is improved by 90% compared with the original LSTM model; and (3) the proposed algorithm can meet the requirements of accurate, real-time, and stable applications of health monitoring. [ABSTRACT FROM AUTHOR]

Copyright of Cotton Textile Technology is the property of Cotton Textile Technology Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

ObjectiveThis paper proposes a lightweight remote sensing ship target detection algorithm LR-YOLO based on improved YOLOv5s to meet the lightweight and fast inference requirements of ship target detection tasks involving remote sensing images. MethodsFirst, the backbone network adopts the ShuffleNet v2 block stacking method, effectively reducing the number of network model parameters and improving the computational speed; second, a region selection module filter is designed to select regions of interest and extract effective features more fully; finally, a circular smooth label is introduced to calculate angle loss and perform rotation detection on remote sensing ship targets, while deformable convolution is used to adapt to geometric deformation and improve detection performance. ResultsThe experimental results on the HRSC2016 ship dataset show that the detection accuracy of the algorithm reaches 92.90%, an improvement of 1.3%, with the number of network model parameters only 39.33% that of the baseline model. ConclusionThe proposed algorithm achieves a balance between lightweight and detection accuracy, providing references for remote sensing ship target detection.

ObjectiveA lightweight and efficient ship detection method based on the YOLO-FNC model is proposed for complex environments such as ports with dense traffic. MethodFirst, a FasterNeXt neural network module is designed on the basis of the FasterNet method and replaces the C3 module in the YOLO model to ensure faster operation without affecting accuracy. Second, a normalization-based attention module (NAM) is integrated into the network structure and the sparse weight penalty is used to suppress the feature weights and ensure more efficient weight calculation. Finally, a new bounding box regression loss is proposed to speed up the prediction frame adjustment and increase the regression rate, thereby improving the convergence rate of the network mode. ResultsThe experimental results show that when performing detection experiments on ship datasets in a self-built complex environment, the proposed method improves the mAP@0.5 by 6.35%, reduces the parameter count by 9.74% and reduces the computational complexity by 11.39%. ConclusionThe proposed method effectively achieves lightweight and high-precision ship detection compared with the YOLOv5s algorithm.

In power line inspections, detecting insulator defects is critical due to the potential breakdown and damage resulting from long-term exposure to the natural environment. However, challenges persist in practical detection processes, such as the limited availability of insulator defect images, difficulty detecting minor defects, and high memory consumption of deep learning models. We propose an insulator defect detection algorithm based on adaptive feature fusion and the lightweight YOLOv5s model to address these issues. First, we enhance the existing dataset by employing techniques like flipping, random pixel manipulation, noise addition, and contrast/brightness adjustments to increase the diversity of training samples. Next, we integrate the adaptive feature fusion (ASFF) module to enable the network to learn relationships between different feature maps, enhancing semantic information and improving the network's ability to detect minor defects. Finally, we replace the backbone of YOLOv5s with the lightweight convolutional neural network EfficientNet, making the network more efficient and enabling it to focus more on target features. Experimental results demonstrate significant improvements in insulator defect detection. The model achieves an accuracy rate of 96.4%, a recall rate of 96.1%, and an mAP of 97.6%, effectively enhancing network performance. [ABSTRACT FROM AUTHOR]

Traffic sign detection and recognition facilitates real-time monitoring and interpretation of various traffic signs on the road, such as those indicating speed limits, prohibition of overtaking, and navigation cues. This has substantial applications for autonomous driving and decision-making systems. Consequently, designing accurate and efficient algorithms for the automatic recognition of traffic signs is crucial in the intelligent transportation field. However, targets that need to be detected by traffic sign recognition applications are mostly small-sized, causing challenges regarding their automatic recognition. The YOLOv5s model, characterized by its minimal depth and narrowest feature map, has gained widespread popularity for executing detection owing to its features of being lightweight and easily portable. Furthermore, the YOLOv5s model uses an anchor-based prediction approach that uses anchor boxes of different sizes and shapes to regress and classify various targets. This method generates dense anchor boxes and enables the model to directly perform object classification and bounding box regression, thereby enhancing its target recall capability. Therefore, the anchor-based Yolov5s method has been applied to traffic sign detection; however, it suffers from issues such as false positives and missed detection. Detection of small targets continues to be a challenging aspect in current traffic sign recognition technology due to the following: small targets carry less information; detection of small targets requires high precision in positioning; and environmental noise may overwhelm the detection of small targets. To overcome the abovementioned issues, such as missed detection, false positives, and low detection accuracy, this study proposes a model called STD-YOLOv5s that is specifically designed for small target detection. First, by increasing the number of upsampling and prediction output layers, this model obtains abundant location information. This can enhance the global understanding of images and solve the issue of insufficient information associated with small targets. Second, the CA attention mechanism is added after each C3 module, whereas the Swin-T attention mechanism module is added before each output layer, increasing the model’s ability to capture multilayer feature information and consequently improving its performance of small target detection. Finally, the accuracy of target localization is ensured using the SIoU penalty function, which considers the target shape and spatial relationships, thereby increasing the model’s ability to capture the positional relationships among targets of different sizes in the image. The STD-YOLOv5s model was validated using the TT100K dataset by ablation and comparison experiments. Experimental results indicate that the proposed model not only maintains the lightweight nature and high detection speed of the YOLOv5s model but also achieves improvements in precision, recall, and average precision.

At present, the combination of edge computing and machine vision has a good application prospect for coal mine safety monitoring. But the storage space and computing resources at the edge are limited, and high-precision complex visual models are difficult to deploy on it. In order to solve the above problems, a YOLOv5s pruning method based on indirect and direct evaluation space fusion (IDESF) is proposed for the edge end of coal mine safety monitoring, aiming to achieve lightweight YOLOv5s network. Firstly, a structural analysis is conducted on the convolutional layers of each module in the YOLOv5s network to determine the free pruning layer and conditional pruning layer. It lays the foundation for subsequent allocation of pruning rates and calculation of the number of pruning kernels. Secondly, the pruning rate is assigned to the prunable layers according to the convolutional kernel weight importance score based on the convolutional kernel weight magnitude and the relative computational complexity of the layers, which effectively reduces the computational complexity of the network after pruning. Thirdly, based on the direct importance evaluation criterion of convolutional kernels, the indirect output importance of convolutional layers is introduced into the direct importance space in the form of scaling factors. The position distribution of convolutional kernels is updated to construct a fused importance evaluation space that includes the output information and amplitude information of convolutional kernels. It thereby improves the comprehensiveness of convolutional kernel importance evaluation. Finally, drawing on the idea of topk voting, the process of median filtering for screening redundant convolution kernels is optimized. The method quantifies the degree of redundancy of a convolutional kernel in terms of the incidence of nodes in the adjacency matrix of a directed graph, which improves the interpretability and generality of the redundant convolutional kernel screening process. The experimental results show the following points. ① From the perspective of balancing model precision and lightweighting, YOLOV5s_IDESF with a pruning rate of 50% is the optimal lightweight YOLOv5s. On the VOC dataset, YOLOv5s_IDESF mAP@.5 and mAP@0.5 is the highest, reaching 0.72 and 0.44 respectively. The parameter count is reduced to a minimum of 2.65×106, the computational complexity is reduced to 1.16×109, and the overall complexity is also reduced to the lowest. The image processing frame rate reaches 31.15 frames per second. ② On the coal mine dataset, YOLOv5s_IDESF mAP@.5 and mAP@0.5∶0.95 achieve the highest values of 0.94 and 0.52, respectively. The parameter count is reduced to a minimum of 3.12×106, the computational complexity is reduced to 1.24×109, and the overall complexity is also minimized. The image processing frame rate reaches 31.55 frames per second.

Abstract Real‐time monitoring of the coal caving process in fully mechanized mining is crucial for achieving intelligent and efficient top‐coal caving. While the coal gangue identification method, employing vision and deep learning, has advanced in the realm of intelligent monitoring, it exhibits a dependency on high‐performance hardware. This reliance poses challenges for deploying identification equipment on mobile terminals, hindering the widespread application of this method. To address the issues above, the paper presents a lightweight algorithm, utilizing You Only Look Once version 5s (YOLOv5s), utilizing YOLOv5s for the real‐time perception of the top‐coal caving state in fully mechanized caving mining. We replace the backbone network of YOLOv5s with the ShuffleNetv2 structure that is more suitable for lightweight deployment, and add the Simple Attention Mechanism attention mechanism to the network structure to enhance the model's receptive field and feature expression ability, and reduce the impact of falling debris on the detection results. A dynamic experimental platform for top‐coal caving in fully mechanized caving mining for thick coal seams is set up, and preprocessing operations such as brightness, sharpening, and denoising are performed on the image data sets collected by high‐speed industrial cameras. Research results show that compared with the traditional YOLOv5s, the improved model's P, mAP, F1 score, and other indicators have increased by 3.4%, 2.1%, and 1.1%, respectively, the model size is 70% of the original, and the detection frames per second value has increased by 48.1%. The lightweight algorithm stabilizes the accuracy of coal gangue identification dramatically in real time. It dramatically reduces the computing pressure on the mobile terminal, providing basic theory and practice for real‐time monitoring of fully mechanized coal caving mining.

Accurate tomato potting seedling classification and identification are crucial for automatic transplanting machines to efficiently carry out potting seedling transplant operations. However, the current potting seedling detection suffers from low precision, inadequate real-time detection, and issues of false detection and omissions. To develop a tomato potting seedling monitoring system ensuring real-time detection and enhancing potting seedling identification accuracy, a tomato potting seedling classification and identification model was proposed based on enhancements to the YOLOv5s target detection algorithm. The research involved creating a tomato potting seedling dataset, incorporating the GAM( global attention mechanism) attention mechanism, implementing the WIOU( wise IoU) loss function strategy, utilizing the deformable convolutional DCN V3, and integrating the CAM(context augmentation module) module. The enhanced model achieved an average detection speed of approximately 12 ms, with the average accuracy AP(average precision) increasing by 3. 8 percentage points compared to the base model, MAP@ 0. 5 rising by 1. 9 percentage points, and R improving by 3. 2 percentage points. When compared under identical experimental conditions, the improved YOLOv5s model exhibits faster detection speed than the commonly used contemporary model, meeting potting seedling detection requirements with enhanced accuracy and improved overall performance, thereby ensuring real-time detection through effective enhancement of tomato potting seedling recognition accuracy. [ABSTRACT FROM AUTHOR]

The traditional detection of rust defects on segmental beams relies on manual identification, which poses high safety risks, low detection efficiency, and unstable accuracy. In order to solve these problems, a segment beam rust detection algorithm based on improved YOLOv5s was proposed, and the attention mechanism CBAM (convolutional block attention module) of channel attention and spatial attention fusion was added to the backbone network module to enhance the ability to extract features. The bi-directional weighted feature pyramid network structure BiFPN (bi-directional feature pyramid networ) was integrated into the neck network module to improve the network’ s detection ability. The efficient intersection over union loss EIoU was combined with Alpha-IoU, which can perform power operation on the loss function, to generate an Alpha-EIoU Loss function to replace the original complete intersection over union loss CIoU, reduce the loss value and further improve the overall performance of the model. After experiments, it has been shown that the improved algorithm outperforms the original YOLOv5s algorithm in terms of accuracy, recall mAP@ 0. 5 and mAP@ 0. 5∶ 0. 95 have increased by 2. 8%, 3. 0%, 2. 0%, and 5. 4% respectively, without adding too many parameters. After the validation of a virtual scene designed by the 3D modeling software 3dsMax, this algorithm can achieve high recognition accuracy in various backgrounds and has strong robustness. It has important theoretical significance and engineering value for the deployment of rust detection in prefabricated segmental beams. [ABSTRACT FROM AUTHOR]

In the process of high-yield cultivation of greenhouse cut flower, high labour consuming and lacking of operational standardization stand out as significant issues. As an important agronomic technique for seedlings rooting, cuttage is a key link in the planting process and an important measure to increase agricultural income, accounting for 70% of the total labor cost. A carnation cuttage robot system was designed to integrate with the grasping-cuttage multifunction for the carnations' automatic cuttage. Taking carnation cuttings as the research object, according to the agronomic requirements under high-yield cultivation models and the materials characteristics of carnation cuttings, an auxiliary cuttage end-effector operation mode and a grasping-cuttage coordinated motion planning model are proposed. It realized high-efficient cuttage operation by the combination of cuttings gripping and auxiliary cuttage. A multi-objective identification and localization model was constructed based on the YOLOv5s deep learning network and the hand-eye coordinate system mapping relationship to meet the needs of target detection. Additionally, a kinematic model was built for the cuttage robot, and an improved MD-H (modified denavit-hartenberg) rule based inverse kinematics modeling method and time optimal trajectory planning algorithm are proposed. A multi-objective cuttage points traversal simulation method was proposed to verify that the inverse kinematics modeling method based on improved MD-H rule can effectively reduce the joint mutation rate and solution time of inverse kinematics. In the framework of Ubuntu 18.04 and ROS-melodic, Moveit-gazebo is used to carry out the joint simulation of time optimal trajectory planning, and the validity and efficiency of time optimal trajectory planning algorithm are verified. Finally, the key technologies of the cuttage robot were integrated to develop the standardized and automated cuttage workflow. We constructed a comprehensive test bench for the cuttage robot to evaluate the localization error and machine performance. The localization average error in X and Y directions during the grasping operation were 2.33 mm and 2.38 mm, the cuttage operation had an average localization error in X, Y and Z directions were 2.21 mm, 2.25 mm and 2.05 mm, respectively. On average, it took 14.787 s in total to operation per plant. The time of recognizing and grasping, cuttage were 6.803 s and 5.104 s, accounting for 46.01% and 34.52% of the total working time respectively. During the test of the performance of cuttage robot, the average qualified rate of cuttage depth and the vertical degree of cuttage were 88% and 92% respectively. The average rate of missing cuttage and root damage were 5% and 2% respectively. The cuttage robot met the basic requirements of standardized cuttage operations. Thus, the study fills gaps in the literature on mechanized cuttage of carnation in China and provides a theoretical basis for standardized seedlings cuttage. The main factors limiting the cuttage efficiency of the robot include: 1) The greater the difference in the shape of the cuttings on the conveyor belt, the longer the detection operation time will be, thus restricting the efficiency of the grasping operation; 2) In order to ensure the safety and accuracy of the robot, it needs to complete two complex actions of grasping and cuttage, so it takes more time to complete the calculation, which affects the overall efficiency of cuttage. The irregular shape of carnation cuttings, the unstable posture and the occlusion of stems and leaves are the main reasons that affected the success rate of visual detection and led to missing cuttage. Future research should be carried out on the attitude estimation algorithm for disordered cuttings to achieve the disordered grasping of multi-objective and multi-pose cuttings. The researches on other crops are needed to further improve the performance and applicability of the robot. Therefore, it can provide more comprehensive theoretical support for the high-efficient and standardized operation of cuttage robot. [ABSTRACT FROM AUTHOR]

The diameter of the optic nerve sheath is an important indicator for assessing the intracranial pressure in critically ill patients. The methods for measuring the optic nerve sheath diameter are generally divided into invasive and non-invasive methods. Compared to the invasive methods, the non-invasive methods are safer and have thus gained popularity. Among the non-invasive methods, using deep learning to process the ultrasound images of the eyes of critically ill patients and promptly output the diameter of the optic nerve sheath offers significant advantages. This paper proposes a CBC-YOLOv5s optic nerve sheath ultrasound image segmentation method that integrates both local and global features. First, it introduces the CBC-Backbone feature extraction network, which consists of dual-layer C3 Swin-Transformer (C3STR) and dual-layer Bottleneck Transformer (BoT3) modules. The C3STR backbone's multi-layer convolution and residual connections focus on the local features of the optic nerve sheath, while the Window Transformer Attention (WTA) mechanism in the C3STR module and the Multi-Head Self-Attention (MHSA) in the BoT3 module enhance the model's understanding of the global features of the optic nerve sheath. The extracted local and global features are fully integrated in the Spatial Pyramid Pooling Fusion (SPPF) module. Additionally, the CBC-Neck feature pyramid is proposed, which includes a single-layer C3STR module and three-layer CReToNeXt (CRTN) module. During upsampling feature fusion, the C3STR module is used to enhance the local and global awareness of the fused features. During downsampling feature fusion, the CRTN module's multi-level residual design helps the network to better capture the global features of the optic nerve sheath within the fused features. The introduction of these modules achieves the thorough integration of the local and global features, enabling the model to efficiently and accurately identify the optic nerve sheath boundaries, even when the ocular ultrasound images are blurry or the boundaries are unclear. The Z2HOSPITAL-5000 dataset collected from Zhejiang University Second Hospital was used for the experiments. Compared to the widely used YOLOv5s and U-Net algorithms, the proposed method shows improved performance on the blurry test set. Specifically, the proposed method achieves precision, recall, and Intersection over Union (IoU) values that are 4.1%, 2.1%, and 4.5% higher than those of YOLOv5s. When compared to U-Net, the precision, recall, and IoU are improved by 9.2%, 21%, and 19.7%, respectively. [ABSTRACT FROM AUTHOR]

In the manufacturing process of cardboard boxes, rapid and accurate surface defect detection is crucial for ensuring quality and preventing resource waste. However, current target detection models have large parameters that impact real-time processing and hinder deployment on devices with limited computational power, reducing both efficiency and applicability. To solve this challenge and better align with real-world application demands, a lightweight improved YOLOv5s model for detecting surface defects on cardboard boxes has been proposed in this study. Inspired by the lightweight features of ShuffleNetV2 and the Ghost model, and with the aim of achieving model lightweighting, the backbone is initially restructured through the incorporation of the ShuffleNetV2 primitive unit. Subsequently, we replace C3 and CBS with C3_Ghost and G_CBS, developing a lightweight backbone and neck structure. Ultimately, inter-layer skip connections are employed to design a feature fusion network applied in the neck to compensate for accuracy. The expriment results demonstrates that the improved YOLOv5-S-G-B model achieves significant reductions in parameters, GFLOPs, and weight dimensions-74.3%, 78.5%, and 72.3%, respectively, compared to the original YOLOv5s. It maintains an average accuracy of 94% and the average detection rate on less powerful CPUs increases from 7.00 frames per second (FPS). The YOLOv5-S-G-B model maintains high accuracy while featuring a significantly lightweight structure, making it better suited for real-world cardboard box defect detection applications. [ABSTRACT FROM AUTHOR]

[Objectives] A light weight YOLOv5S crop pest target detection model was proposed to solve the problem of identifying crop pests, when the sample quantity was insufficient. [Methods] Two Ghost Bottle Block linear feature extraction modules were encapsulated into one GB module using Ghost technology, replacing the first seven CBL, CSP and SPP nonlinear feature extraction modules in YOLOv5S, which reduced the parameters and the volume of YOLOv5S. [Results] The network's dependence on computing hardware and training samples was reduced while ensuring the effectiveness of pest detection. In order to verify the effectiveness of the model, 12 types of pests were identified and located for 5 crops, including rice, corn, cotton, potatoes and alfalfa. The multi category mean average precision (mAP) was 91.31%, which was 2.5 percentage points higher than the YOLOv5S model. [Conclusions] By comparing with SSD, Faster RCNN and YOLOv5S, it was found that the model proposed in this paper had advantages in mAP, F1 score, precision and recall evaluation indicators, especially in the detection of small targets, dense targets and targets similar to the background. [ABSTRACT FROM AUTHOR]

[Objectives] In order to achieve precise herbicide application in lawn maintenance, this study addresses the challenge of weed identification due to the similar colors of weeds and turfgrass in natural environments. Based on the YOLOv5s target detection network, a model highlighting the color information of input images was proposed for lawn weed detection. [Methods]The input images were sliced, and the RGB channels were expanded to four times their original number. The sliced images underwent global average pooling and global max pooling to extract color information from the original images. The color channels were weighted to enhance the detection of monochromatic targets. To further improve the network's ability to recognize targets with similar shapes but different colors, the SPPF module was modified to the SPPFCSPC module, which increased the accuracy of weed identification in various lawn environments. [Results] The improved YOLOv5s network showed a good recognition effect on monochromatic targets, achieving a detection accuracy of 0.928 for common lawn weeds, a recall rate of 0.970, an mAP@0.5 of 0.943, and a frame rate of 104 f·s-1, with a model memory usage of 13.8 MB. The detection accuracy improved by 4.1 percentage points compared to YOLOv5s using the Focus module and by 5.4 percentage points compared to YOLOv5s using the Conv module. The detection results in sparse woodland lawns achieved an mAP@0.5 of 0.957. [Conclusions] The algorithm effectively identified various lawn weeds in natural environments and supported precise herbicide application for lawn weeds. The improved module demonstrated good sensitivity to color, providing a technical reference for the detection of monochromatic targets with high background color similarity. [ABSTRACT FROM AUTHOR]

Currently, multi-layer deep convolutional networks are mostly used for field weed recognition to extract and identify target features. However, in practical application scenarios, they still face challenges such as insufficient recognition accuracy, a large number of model parameters, and slow detection speed. In response to the above problems, using chickweed as the identification object, a weed identification model based on improved YOLOv5s was proposed. Firstly, the Squeeze-and-Excitation Module (SE) and Convolutional Block Attention Module (CBAM) were added to the model's feature extraction network to improve the model's recognition accuracy; secondly, the Ghost convolution lightweight feature fusion network was introduced to effectively identify the volume, parameter amount, and calculation amount of the model, and make the model lightweight; finally, we replaced the loss function in the original target bounding box with the Efficient Intersection over Union (EloU) loss function to further improve the detection performance of the improved YOLOv5s model. After testing, the accuracy of the improved YOLOv5s model was 96.80%, the recall rate was 94.00%, the average precision was 93.20%, and the frame rate was 14.01 fps, which were improved by 6.6%, 4.4%, 1.0%, and 6.1%, respectively, compared to the original YOLOv5s model. The model volume was 9.6 MB, the calculation amount was 13.6 GB, and the parameter amount was 5.9 MB, which decreased by 29.4%, 14.5%, and 13.2% compared with the original YOLOv5s model, respectively. This model can effectively distinguish chickweed between crops. This research can provide theoretical and technical support for efficient identification of weeds in complex field environments. [ABSTRACT FROM AUTHOR]

Wheat spike count is one of the critical indicators for assessing the growth and yield of wheat. However, illumination variations, mutual occlusion, and background interference have greatly affected wheat spike detection. A lightweight detection method was proposed based on the YOLOv5s. Initially, the original YOLOv5s was improved by combing the additional small-scale detection layer and integrating the ECA (Efficient Channel Attention) attention mechanism into all C3 modules (YOLOv5s + 4 + ECAC3). After comparing GhostNet, ShuffleNetV2, and MobileNetV3, the GhostNet architecture was finally selected as the optimal lightweight model framework based on its superior performance in various evaluations. Subsequently, the incorporation of five different activation functions into the network led to the identification of the RReLU (Randomized Leaky ReLU) activation function as the most effective in augmenting the network's performance. Ultimately, the network's loss function of CIoU (Complete Intersection over Union) was optimized using the EIoU (Efficient Intersection over Union) loss function. Despite a minor reduction of 2.17% in accuracy for the refined YOLOv5s + 4 + ECAC3 + G + RR + E network when compared to the YOLOv5s + 4 + ECAC3, there was a marginal improvement of 0.77% over the original YOLOv5s. Furthermore, the parameter count was diminished by 32% and 28.2% relative to the YOLOv5s + 4 + ECAC3 and YOLOv5s, respectively. The model size was reduced by 28.0% and 20%, and the Giga Floating-point Operations Per Second (GFLOPs) were lowered by 33.2% and 9.5%, respectively, signifying a substantial improvement in the network's efficiency without significantly compromising accuracy. This study offers a methodological reference for the rapid and accurate detection of agricultural objects through the enhancement of a deep learning network. [ABSTRACT FROM AUTHOR]

Road crack detection is of paramount importance for ensuring vehicular traffic safety, and implementing traditional detection methods for cracks inevitably impedes the optimal functioning of traffic. In light of the above, we propose a USSC-YOLO-based target detection algorithm for unmanned aerial vehicle (UAV) road cracks based on machine vision. The algorithm aims to achieve the high-precision detection of road cracks at all scale levels. Compared with the original YOLOv5s, the main improvements to USSC-YOLO are the ShuffleNet V2 block, the coordinate attention (CA) mechanism, and the Swin Transformer. First, to address the problem of large network computational spending, we replace the backbone network of YOLOv5s with ShuffleNet V2 blocks, reducing computational overhead significantly. Next, to reduce the problems caused by the complex background interference, we introduce the CA attention mechanism into the backbone network, which reduces the missed and false detection rate. Finally, we integrate the Swin Transformer block at the end of the neck to enhance the detection accuracy for small target cracks. Experimental results on our self-constructed UAV near–far scene road crack i(UNFSRCI) dataset demonstrate that our model reduces the giga floating-point operations per second (GFLOPs) compared to YOLOv5s while achieving a 6.3% increase in mAP@50 and a 12% improvement in mAP@ [50:95]. This indicates that the model remains lightweight meanwhile providing excellent detection performance. In future work, we will assess road safety conditions based on these detection results to prioritize maintenance sequences for crack targets and facilitate further intelligent management. [ABSTRACT FROM AUTHOR]

As the foundation of circuits and electronic systems, the quality of printed circuit boards (PCBs) is directly related to the reliability of the whole system. At present, deep learning-based target detection technology has a broad application prospect in PCB defect detection. However, there are still some shortcomings in the process of detection. For example, the detection speed is slow, and the detection accuracy is not ideal for defects with small size. Therefore, this paper proposes an improved algorithm based on the YOLOv5s model to improve the detection of PCB defects. The algorithm enhances the connection between the shallow and high layers of the network by improving the backbone network structure. Meanwhile, a self-attention mechanism is introduced into YoloHead to enhance the feature extraction and the ability to focus on the global information of the feature map, and the self-attention mechanism is optimized by combining the computational characteristics of the SENet, the CBAM and the residual unit. While improving the accuracy of PCB defect detection, better computational efficiency is ensured. In subsequent experiments, this paper trains and tests the improved YOLOv5s algorithm using publicly available PCB defect datasets. The results show that the algorithm achieves a detection rate of 39.7 frames per second and a maximum average precision of 96.14%, which is a 4.68% improvement over the original YOLOv5s. [ABSTRACT FROM AUTHOR]

Summary: During harvesting, transportation and storage of kiwifruit, the flesh is often bruised by collision or compression. However, the bruises in kiwifruit are extremely difficult to recognise by naked eyes and are called hidden bruises. Accordingly, a fast method for detecting hidden bruises in kiwifruit was developed in this study based on hyperspectral imaging (HSI) coupled with deep learning. The spectral range (924–1277 nm) and feature wavelengths (928.19, 1051.03 and 1190.47 nm) sensitive to hidden bruises in kiwifruit were selected using the principal component analysis (PCA). Subsequently, three‐channel images (Dataset 1), grayscale images (Dataset 2) and pseudo‐colour images (Dataset 3) were generated according to the images of feature wavelengths of the kiwifruit. The YOLOv5s model for detecting the hidden bruised areas of the kiwifruit was developed using these three datasets. The results showed that the YOLOv5s detection model performed best at Dataset 1, and the values of Precision, Recall, F1, mAP and FNR of this model were 98.25%, 97.50%, 97.87%, 99.12% and 2.50% respectively. The study showed that HSI technology combined with the YOLOv5s model can effectively detect hidden bruises in kiwifruit, providing references for detecting hidden bruises in other fruit. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Henan University of Technology Natural Science Edition is the property of Henan University of Technology Journal Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The growth environment of green walnuts is complex. In the actual picking and identification process, interference from near-background colors, occlusion by branches and leaves, and excessive model complexity pose higher demands on the performance of walnut detection algorithms. Therefore, a lightweight walnut detection algorithm suitable for complex environments is proposed based on YOLOv5s. First, the backbone network is reconstructed using the lightweight GhostNet network, laying the foundation for a lightweight model architecture. Next, the C3 structure in the feature fusion layer is optimized by proposing a lightweight C3 structure to enhance the model's focus on important walnut features. Finally, the loss function is improved to address the problems of target loss and gradient adaptability during training. To further reduce model complexity, the improved algorithm undergoes pruning and knowledge distillation operations, and is then deployed and tested on small edge devices. Experimental results show that compared to the original YOLOv5s model, the improved algorithm reduces the number of parameters by 72.9% and the amount of computation by 84.1%. The mAP0.5 increased by 1.1%, the precision increased by 0.7%, the recall increased by 0.3%, and the FPS is 179.6% of the original model, meeting the real-time detection needs for walnut recognition and providing a reference for walnut harvesting identification. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Henan Agricultural Sciences is the property of Editorial Board of Journal of Henan Agricultural Sciences and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In low-light environments, the presence of numerous small, dense, and occluded objects challenges the effectiveness of conventional object detection methods, failing to achieve desirable results. To address this, this paper proposes an efficient object detection network, YOLO_GD, which is designed for precise detection of targets in low-light scenarios. This algorithm, based on the foundational framework of YOLOv5s, implements a cross-layer feature fusion method founded on an information gathering and distribution mechanism. This method mitigates the issue of information loss during inter-layer feature exchange and, building on this, constructs a Bi-level routing spatial attention module to reduce computational redundancy caused by the self-attention mechanism, thereby enhancing the model's detection accuracy for small objects. Furthermore, through the introduction of a novel deformable convolution, a cross-stage local feature fusion module is established, enabling the model to capture the complex features of input data more accurately and improve detection precision for dense objects. Lastly, the introduction of a probabilistic distance metric in the bounding box regression loss function enhances the network model's generalization capability, further increasing detection accuracy in occluded scenarios. Experimental results on the ExDark dataset demonstrate that compared to YOLOv5, there is a 5.97% improvement in mean average precision (mAP), effectively enhancing object detection performance in low-light conditions. [ABSTRACT FROM AUTHOR]

To tackle the issue of low detection accuracy in insulator images caused by intricate backgrounds and small defect sizes, as well as the requirement for real-time detection on embedded and mobile devices, this research introduces the YOLOv5s-KE model. Integrating multiple strategies, YOLOv5s-KE aims to boost detection accuracy significantly. Initially, an enhanced anchor generation method utilizing the K-means++ algorithm is proposed to generate more appropriate anchor boxes for insulator defects. Moreover, an attention mechanism is integrated into both the backbone and neck networks to enhance the model's capacity to focus on defect features and resist interference. To improve the detection of small defects, the EIoU loss function is implemented in place of the original CIoU loss function. In order to meet the real-time detection needs on embedded and mobile devices, the model is further refined through the integration of Ghost convolution for lightweight feature extraction and a linear transformation to reduce the computational burden of standard convolution. A channel pruning strategy is deployed to optimize the sparsely trained network, diminishing redundancy, and improving model generalization. Additionally, the CARAFE operator replaces the original upsampling operator to minimize model parameters and elevate detection speed. Experimental outcomes demonstrate that YOLOv5s-KE achieves a detection accuracy of 92.3% on the Chinese transmission line insulator dataset, marking a 5.2% enhancement over the original YOLOv5s. The streamlined version of YOLOv5s-KE achieves a detection speed of 94.3 frames per second, indicating an improvement of 30.1 frames per second compared to the original model. Model parameters are condensed to 9.6 M, resulting in a detection accuracy of 91.1%. This study underscores the precision and efficiency of the proposed approach, suggesting that the advanced strategies explored introduce novel possibilities for insulator defect detection. [ABSTRACT FROM AUTHOR]

Copyright of Computer Measurement & Control is the property of Magazine Agency of Computer Measurement & Control and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The excessive use of electronic devices for prolonged periods has led to problems such as neck pain and pressure injury in sedentary people. If not detected and corrected early, these issues can cause serious risks to physical health. Detectors for generic objects cannot adequately capture such subtle neck behaviors, resulting in missed detections. In this paper, we explore a deep learning-based solution for detecting abnormal behavior of the neck and propose a model called NABNet that combines object detection based on YOLOv5s with pose estimation based on Lightweight OpenPose. NABNet extracts the detailed behavior characteristics of the neck from global to local and detects abnormal behavior by analyzing the angle of the data. We deployed NABNet on the cloud and edge devices to achieve remote monitoring and abnormal behavior alarms. Finally, we applied the resulting NABNet-based IoT system for abnormal behavior detection in order to evaluate its effectiveness. The experimental results show that our system can effectively detect abnormal neck behavior and raise alarms on the cloud platform, with the highest accuracy reaching 94.13%. [ABSTRACT FROM AUTHOR]

Modern aquaculture utilizes computer vision technology to analyze underwater images of fish, contributing to optimized water quality and improved production efficiency. The purpose of this study is to efficiently perform underwater fish detection and tracking using multi-object tracking (MOT) technology. To achieve this, the FairMOT model was employed to simultaneously implement pixel-level object detection and re-identification (Re-ID) functions, comparing two backbone models: FairMOT+YOLOv5s and FairMOT+DLA-34. The study constructed a dataset targeting the popular black porgy in Korean aquaculture, using underwater video data from five different environments collected from the internet. During the training process, the FairMOT+YOLOv5s model rapidly reduced train loss and demonstrated stable performance. The FairMOT+DLA-34 model showed better results in ID tracking performance, with an accuracy of 44.1%, an IDF1 of 11.0%, an MOTP of 0.393, and an IDSW of 1. In contrast, the FairMOT+YOLOv5s model recorded an accuracy of 43.8%, an IDF1 of 14.6%, an MOTP of 0.400, and an IDSW of 10. The results of this study indicate that the FairMOT+YOLOv5s model demonstrated higher IDF1 and MOTP scores compared to the FairMOT+DLA-34 model, while the FairMOT+DLA-34 model showed superior performance in ID tracking accuracy and had fewer ID switches. [ABSTRACT FROM AUTHOR]

In order to improve the poor accuracy of steel surface defects caused by poor detection of small targets, on the basis of the YOLOv5s, by adding the SE (attention mechanism) in the backbone network mechanism, C2f module instead of C3 module, the BIFPN (bidirectional feature pyramid network) instead of the PAN (path aggregation network) network in the neck,these three methods were used to investigate the improvement of the ability to the defect of small target detection. It aims to improve the detection accuracy and achieve real-time detection. The results show that the mAP (mean average precision) of the improved YOLOv5s-SCB algorithm on NEU-DET (northeastern university-detect) reaches 77. 9%, which is 3. 7% higher than that of the YOLOv5s network on the premise of real-time detection. Compared with other improved algorithms based on YOLOv5s and YOLOv8, YOLOv5S-SCB achieves better detection effect. It is concluded that the proposed steel surface defect detection algorithm YOLOv5S-SCB can better detect defects on steel surfaces. [ABSTRACT FROM AUTHOR]

To achieve efficient and real-time classification and recognition of freshwater fish on limited equipment resources, this study proposes a method based on the YOLOv5s network to learn freshwater fish dataset images based on machine vision technology. Firstly, replace the backbone network of YOLOv5s with ShuffleNetV2 of the lightweight network to improve the efficiency of architecture design and ensure the detection speed. Secondly, the lightweight convolution GSConv is introduced to reduce the complexity of the network, maintain the accuracy of the network, and achieve the lightweight effect. Finally, the combination of ShuffleNetV2 and lightweight convolution GSConv can significantly improve the lightweight effect while maintaining accuracy. The algorithm was validated with four datasets of Rhodeinae, goldfish, grass carp, and stone moroko, which are all typical carp. The results showed that the improved network model FLOPs is 4.9 G, which was 69% less than YOLOv5s, the model size was 6.1 MB, which was 58% less than YOLOv5s, and the improved model mAP0.5 could reach 95.5%. [ABSTRACT FROM AUTHOR]

The shape and quantity of Lentinus edodes (commonly known as shiitake) fruiting bodies significantly affect their quality and yield. Accurate and rapid segmentation of these fruiting bodies is crucial for quality grading and yield prediction. This study proposed the YOLOv5seg-BotNet, a model for the instance segmentation of Lentinus edodes, to research its application for the mushroom industry. First, the backbone network was replaced with the BoTNet, and the spatial convolutions in the local backbone network were replaced with global self-attention modules to enhance the feature extraction ability. Subsequently, the PANet was adopted to effectively manage and integrate Lentinus edodes images in complex backgrounds at various scales. Finally, the Varifocal Loss function was employed to adjust the weights of different samples, addressing the issues of missed segmentation and mis-segmentation. The enhanced model demonstrated improvements in the precision, recall, Mask_AP, F1-Score, and FPS, achieving 97.58%, 95.74%, 95.90%, 96.65%, and 32.86 frames per second, respectively. These values represented the increases of 2.37%, 4.55%, 4.56%, 3.50%, and 2.61% compared to the original model. The model achieved dual improvements in segmentation accuracy and speed, exhibiting excellent detection and segmentation performance on Lentinus edodes fruiting bodies. This study provided technical fundamentals for future application of image detection and decision-making processes to evaluate mushroom production, including quality grading and intelligent harvesting. [ABSTRACT FROM AUTHOR]