Search

The thermal efficiency for the pelletizing process is intrinsically linked to the diameter and humidity of the iron ore pellets, so that the sensing of the granulometric range in the formation of the pellets becomes essential to the flow of the pelletizing process in the steel industry; this paper presents the assembly of a computer vision system for the detection and segmentation of pellets aiming at the automation of the granulometric measurement, following its formation in the pelletizing disk, using the instance segmentation method to verify whether the particle granulometric distribution (PSD) is adequate for "real-time" applications. The system calculates the normal distribution of the diameter in millimeters, evaluating the normal curve and the standard deviation of the segmented pellets, using a deep neural network based on the You Only Look At CoefficienTs (YOLACT) network, adding speed and precision in the granulometric analysis. In the sample sets, the need for adjustment factors inherent to the pelletizing process became evident. This led to the establishment of the computer vision system, termed the Volumetric Correction Factor (VCF) and Visual Overlay Factor (VOF). The VCF is utilized to estimate the volume of pellets within the pelletizing disk during operation, while the VOF adjusts the millimeter-per-pixel (mpp) ratio. The results of the measurement system proved to be efficient in real-time granulometric measurement. [ABSTRACT FROM AUTHOR]

ObjectiveCamellia oleifera is one of the four major woody oil plants in the world. Diseases is a significant factor leading to the decline in quality of Camellia oleifera and the financial loss of farmers. Among these diseases, anthracnose is a common and severe disease in Camellia oleifera forests, directly impacting yields and production rates. Accurate disease assessment can improve the prevention and control efficiency and safeguarding the farmers' profit. In this study, an improved You Only Look at CoefficienTs (YOLACT) based method was proposed to realize automatic and efficient grading of the severity of Camellia oleifera leaf anthracnose.MethodsHigh-resolution images of Camellia oleifera anthracnose leaves were collected using a smartphone at the National Camellia oleifera Seed Base of Jiangxi Academy of Forestry, and finally 975 valid images were retained after a rigorous screening process. Five data enhancement means were applied, and a data set of 5 850 images was constructed finally, which was divided into training, validation, and test sets in a ratio of 7:2:1. For model selection, the Camellia-YOLACT model was proposed based on the YOLACT instance segmentation model, and by introducing improvements such as Swin-Transformer, weighted bi-directional feature pyramid network, and HardSwish activation function. The Swin Transformer was utilized for feature extraction in the backbone network part of YOLACT, leveraging the global receptive field and shift window properties of the self-attention mechanism in the Transformer architecture to enhance feature extraction capabilities. Additionally, a weighted bidirectional feature pyramid network was introduced to fuse feature information from different scales to improve the detection ability of the model for objects at different scales, thereby improving the detection accuracy. Furthermore, to increase the the model's robustness against the noise in the input data, the HardSwish activation function with stronger nonlinear capability was adopted to replace the ReLu activation function of the original model. Since images in natural environments usually have complex background and foreground information, the robustness of HardSwish helped the model better handling these situations and further improving the detection accuracy. With the above improvements, the Camellia-YOLACT model was constructed and experimentally validated by testing the Camellia oleifera anthracnose leaf image dataset.Results and DiscussionsA transfer learning approach was used for experimental validation on the Camellia oleifera anthracnose severity grading dataset, and the results of the ablation experiments showed that the mAP75 of Camellia-YOLACT proposed in this study was 86.8%, mAPall was 78.3%, mAR was 91.6% which were 5.7%, 2.5% and 7.9% higher than YOLACT model. In the comparison experiments, Camellia-YOLACT performed better than Segmenting Objects by Locations (SOLO) in terms of both accuracy and speed, and its detection speed was doubled compared to Mask R-CNN algorithm. Therefore, the Camellia-YOLACT algorithm was suitable in Camellia oleifera gardens for anthracnose real-time segmentation. In order to verify the outdoors detection performance of Camellia-YOLACT model, 36 groups of Camellia oleifera anthracnose grading experiments were conducted. Experimental results showed that the grading correctness of Camellia oleifera anthracnose injection severity reached 94.4%, and the average absolute error of K-value was 1.09%. Therefore, the Camellia-YOLACT model proposed in this study has a better performance on the grading of the severity of Camellia oleifera anthracnose.ConclusionsThe Camellia-YOLACT model proposed got high accuracy in leaf and anthracnose segmentation of Camellia oleifera, on the basis of which it can realize automatic grading of the severity of Camellia oleifera anthracnose. This research could provide technical support for the precise control of Camellia oleifera diseases.

[Objective] Camellia oleifera is one of the four major woody oil plants in the world. Diseases is a significant factor leading to the de‐ cline in quality of Camellia oleifera and the financial loss of farmers. Among these diseases, anthracnose is a common and severe dis‐ ease in Camellia oleifera forests, directly impacting yields and production rates. Accurate disease assessment can improve the preven‐ tion and control efficiency and safeguarding the farmers' profit. In this study, an improved You Only Look at CoefficienTs (YOLACT) based method was proposed to realize automatic and efficient grading of the severity of Camellia oleifera leaf anthracnose. [Methods] High-resolution images of Camellia oleifera anthracnose leaves were collected using a smartphone at the National Camellia oleifera Seed Base of Jiangxi Academy of Forestry, and finally 975 valid images were retained after a rigorous screening process. Five data enhancement means were applied, and a data set of 5 850 images was constructed finally, which was divided into training, validation, and test sets in a ratio of 7:2:1. For model selection, the Camellia-YOLACT model was proposed based on the YOLACT instance segmentation model, and by introducing improvements such as Swin-Transformer, weighted bi-directional feature pyramid network, and HardSwish activation function. The Swin Transformer was utilized for feature extraction in the backbone network part of YOLACT, leveraging the global receptive field and shift window properties of the self-attention mechanism in the Transformer archi‐ tecture to enhance feature extraction capabilities. Additionally, a weighted bidirectional feature pyramid network was introduced to fuse feature information from different scales to improve the detection ability of the model for objects at different scales, thereby im‐ proving the detection accuracy. Furthermore, to increase the the model's robustness against the noise in the input data, the HardSwish activation function with stronger nonlinear capability was adopted to replace the ReLu activation function of the original model. Since images in natural environments usually have complex background and foreground information, the robustness of HardSwish helped the model better handling these situations and further improving the detection accuracy. With the above improvements, the CamelliaYOLACT model was constructed and experimentally validated by testing the Camellia oleifera anthracnose leaf image dataset. [Results and Discussions] A transfer learning approach was used for experimental validation on the Camellia oleifera anthracnose severity grading dataset, and the results of the ablation experiments showed that the mAP75 of Camellia-YOLACT proposed in this study was 86.8%, mAPall was 78.3%, mAR was 91.6% which were 5.7%, 2.5% and 7.9% higher than YOLACT model. In the comparison experiments, Camellia-YOLACT performed better than Segmenting Objects by Locations (SOLO) in terms of both accuracy and speed, and its detection speed was doubled compared to Mask R-CNN algorithm. Therefore, the Camellia-YOLACT algorithm was suitable in Camellia oleifera gardens for anthracnose real-time segmentation. In order to verify the outdoors detection performance of Camellia-YOLACT model, 36 groups of Camellia oleifera anthracnose grading experiments were conducted. Experimental results showed that the grading correctness of Camellia oleifera anthracnose injection severity reached 94.4%, and the average absolute error of K-value was 1.09%. Therefore, the Camellia-YOLACT model proposed in this study has a better performance on the grading of the severity of Camellia oleifera anthracnose. [Conclusions] The Camellia-YOLACT model proposed got high accuracy in leaf and anthracnose segmentation of Camellia oleifera, on the basis of which it can realize automatic grading of the severity of Camellia oleifera anthracnose. This research could provide technical support for the precise control of Camellia oleifera diseases. [ABSTRACT FROM AUTHOR]

Among the existing insulator defect detection methods, the automatic detection of inspection robots based on the instance segmentation algorithm is relatively more efficient, but the problem of the limited accuracy of the segmentation algorithm is still a bottleneck for increasing inspection efficiency. Therefore, we propose a single-stage insulator instance defect segmentation method based on both an attention mechanism and improved feature fusion network. YOLACT is selected as the basic instance segmentation model. Firstly, to improve the segmentation speed, MobileNetV2 embedded with an scSE attention mechanism is introduced as the backbone network. Secondly, a new feature map that combines semantic and positional information is obtained by improving the FPN module and fusing the feature maps of each layer, during which, an attention mechanism is introduced to further improve the quality of the feature map. Thirdly, in view of the problems that affect the insulator segmentation, a Restrained-IoU (RIoU) bounding box loss function which covers the area deviation, center deviation, and shape deviation is designed for object detection. Finally, for the validity evaluation of the proposed method, experiments are performed on the insulator defect data set. It is shown in the results that the improved algorithm achieves a mask accuracy improvement of 5.82% and a detection speed of 37.4 FPS, which better complete the instance segmentation of insulator defect images. [ABSTRACT FROM AUTHOR]

A garment image instance segmentation method based on improved YOLACT was proposed to solve the problem of low accuracy and speed of clothing image instance segmentation. Based on the YOLACT model, firstly, the depth separable convolution was used in the ResNetlOl network to replace the traditional convolution, reduce the amount of model calculation and model parameters, and accelerate the speed of the model. Then, the efficient channel attention module was introduced to optimize the output features after the protonet, capture the cross-channel interaction information of the clothing image, and strengthen the feature extraction ability of mask branches. Finally, the Leaky ReLU activation function was used in the training process to ensure that the weight information is updated in time, and to improve the model's ability to extract the negative feature information of the clothing image. The experimental results show that compared with the original model, the proposed method can effectively reduce the number of model parameters, and increase the accuracy and the speed. The speed increased by 4. 82 frame per second, and the average accuracy increased by 5. 4%. [ABSTRACT FROM AUTHOR]

Abstract Distributed deep learning is a promising approach for training and deploying large and complex deep learning models. This paper presents a comprehensive workflow for deploying and optimizing the YOLACT instance segmentation model as on big data clusters. OpenVINO, a toolkit known for its high-speed data processing and ability to optimize deep learning models for deployment on a variety of devices, was used to optimize the YOLACT model. The model is then run on a big data cluster using BigDL, a distributed deep learning library for Apache Spark. BigDL provides a high-level programming interface for defining and training deep neural networks, making it suitable for large-scale deep learning applications. In distributed deep learning, input data is divided and distributed across multiple machines for parallel processing. This approach offers several advantages, including the ability to handle very large data that can be stored in a distributed manner, scalability to decrease processing time by increasing the number of workers, and fault tolerance. The proposed workflow was evaluated on virtual machines and Azure Databricks, a cloud-based platform for big data analytics. The results indicated that the workflow can scale to large datasets and deliver high performance on Azure Databricks. This study explores the benefits and challenges of using distributed deep learning on big data clusters for instance segmentation. Popular distributed deep learning frameworks are discussed, and BigDL is chosen. Overall, this study highlights the practicality of distributed deep learning for deploying and scaling sophisticated deep learning models on big data clusters.

Distributed deep learning is a promising approach for training and deploying large and complex deep learning models. This paper presents a comprehensive workflow for deploying and optimizing the YOLACT instance segmentation model as on big data clusters. OpenVINO, a toolkit known for its high-speed data processing and ability to optimize deep learning models for deployment on a variety of devices, was used to optimize the YOLACT model. The model is then run on a big data cluster using BigDL, a distributed deep learning library for Apache Spark. BigDL provides a high-level programming interface for defining and training deep neural networks, making it suitable for large-scale deep learning applications. In distributed deep learning, input data is divided and distributed across multiple machines for parallel processing. This approach offers several advantages, including the ability to handle very large data that can be stored in a distributed manner, scalability to decrease processing time by increasing the number of workers, and fault tolerance. The proposed workflow was evaluated on virtual machines and Azure Databricks, a cloud-based platform for big data analytics. The results indicated that the workflow can scale to large datasets and deliver high performance on Azure Databricks. This study explores the benefits and challenges of using distributed deep learning on big data clusters for instance segmentation. Popular distributed deep learning frameworks are discussed, and BigDL is chosen. Overall, this study highlights the practicality of distributed deep learning for deploying and scaling sophisticated deep learning models on big data clusters. [ABSTRACT FROM AUTHOR]

The real-time detection and instance segmentation of strawberries constitute fundamental components in the development of strawberry harvesting robots. Real-time identification of strawberries in an unstructured environment is a challenging task. Current instance segmentation algorithms for strawberries suffer from issues such as poor real-time performance and low accuracy. To this end, the present study proposes an Efficient YOLACT (E-YOLACT) algorithm for strawberry detection and segmentation based on the YOLACT framework. The key enhancements of the E-YOLACT encompass the development of a lightweight attention mechanism, pyramid squeeze shuffle attention (PSSA), for efficient feature extraction. Additionally, an attention-guided context-feature pyramid network (AC-FPN) is employed instead of FPN to optimize the architecture's performance. Furthermore, a feature-enhanced model (FEM) is introduced to enhance the prediction head's capabilities, while efficient fast non-maximumsuppression (EF-NMS) is devised to improve non-maximumsuppression. The experimental results demonstrate that the E-YOLACT achieves a Box-mAP andMask-mAP of 77.9 and 76.6, respectively, on the custom dataset.Moreover, it exhibits an impressive category accuracy of 93.5%. Notably, the E-YOLACT also demonstrates a remarkable real-time detection capability with a speed of 34.8 FPS. The method proposed in this article presents an efficient approach for the vision system of a strawberry-picking robot. [ABSTRACT FROM AUTHOR]

Autonomous driving technology has made a lot of outstanding achievements with deep learning, and the vehicle detection and classification algorithm has become one of the critical technologies of autonomous driving systems. The vehicle instance segmentation can perform instance-level semantic parsing of vehicle information, which is more accurate and reliable than object detection. However, the existing instance segmentation algorithms still have the problems of poor mask prediction accuracy and low detection speed. Therefore, this paper proposes an advanced real-time instance segmentation model named FIR-YOLACT, which fuses the ICIoU (Improved Complete Intersection over Union) and Res2Net for the YOLACT algorithm. Specifically, the ICIoU function can effectively solve the degradation problem of the original CIoU loss function, and improve the training convergence speed and detection accuracy. The Res2Net module fused with the ECA (Efficient Channel Attention) Net is added to the model’s backbone network, which improves the multi-scale detection capability and mask prediction accuracy. Furthermore, the Cluster NMS (Non-Maximum Suppression) algorithm is introduced in the model’s bounding box regression to enhance the performance of detecting similarly occluded objects. The experimental results demonstrate the superiority of FIR-YOLACT to the based methods and the effectiveness of all components. The processing speed reaches 28 FPS, which meets the demands of real-time vehicle instance segmentation. [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.)

This study develops an efficient approach for precise channel frame detection in complex backgrounds, addressing the critical need for accurate drone navigation. Leveraging YOLACT and group regression, our method outperforms conventional techniques that rely solely on color information. We conducted extensive experiments involving channel frames placed at various angles and within intricate backgrounds, training the algorithm to effectively recognize them. The process involves initial edge image detection, noise reduction through binarization and erosion, segmentation of channel frame line segments using the Hough Transform algorithm, and subsequent classification via the K-means algorithm. Ultimately, we obtain the regression line segment through linear regression, enabling precise positioning by identifying intersection points. Experimental validations validate the robustness of our approach across diverse angles and challenging backgrounds, making significant advancements in UAV applications.

Thousands of tonnes of litter enter the ocean every day, posing a significant threat to marine life and ecosystems. While floating and beach litter are often in the spotlight, about 70% of marine litter eventually sinks to the seafloor, making underwater litter the largest accumulation of marine litter that often goes undetected. Plastic debris makes up the majority of ocean litter and is a known source of microplastics in the ocean. This paper focuses on the detection of ocean plastic using neural network models. Two neural network models will be trained, i.e., YOLACT and the Mask R-CNN, for the instance segmentation of underwater litter in images. The models are trained on the TrashCAN dataset, using pre-trained model weights trained using COCO. The trained neural network could achieve a mean average precision ( m A P ) of 0.377 and 0.365 for the Mask R-CNN and YOLACT, respectively. The lightweight nature of YOLACT allows it to detect images at up to six times the speed of the Mask R-CNN, while only making a comparatively smaller trade-off in terms of performance. This allows for two separate applications: YOLACT for the collection of litter using autonomous underwater vehicles (AUVs) and the Mask R-CNN for surveying litter distribution. [ABSTRACT FROM AUTHOR]

After years of service, bridges could lose their expected functions. Considering the significant number of bridges and the adverse inspecting environment, the urgent requirement for timely and efficient inspection solutions, such as computer vision techniques, have been attractive in recent years, especially for those bridge components with poor accessibility. In this paper, a lightweight procedure for bridge apparent-defect detection is proposed, including a crack annotation method and crack detection. First of all, in order to save computational costs and improve generalization performance, we propose herein a relative-breakpoint annotation method to build a crack instance segmentation dataset, a critical process for a supervised vision-based crack detection method. Then, the trained models based on classic Mask RCNN and Yolact are transferred to evaluate the effectiveness of the proposed method. To verify the correctness, universality and generality of the proposed crack-detection framework, approximately 800 images are used for model training, while nearly 100 images are saved for validation. Results show that the crack instance segmentation model can achieve a level of 90% for both accuracy and recall values, with a limited dataset. [ABSTRACT FROM AUTHOR]

Among the existing insulator defect detection methods, the automatic detection of inspection robots based on the instance segmentation algorithm is relatively more efficient, but the problem of the limited accuracy of the segmentation algorithm is still a bottleneck for increasing inspection efficiency. Therefore, we propose a single-stage insulator instance defect segmentation method based on both an attention mechanism and improved feature fusion network. YOLACT is selected as the basic instance segmentation model. Firstly, to improve the segmentation speed, MobileNetV2 embedded with an scSE attention mechanism is introduced as the backbone network. Secondly, a new feature map that combines semantic and positional information is obtained by improving the FPN module and fusing the feature maps of each layer, during which, an attention mechanism is introduced to further improve the quality of the feature map. Thirdly, in view of the problems that affect the insulator segmentation, a Restrained-IoU (RIoU) bounding box loss function which covers the area deviation, center deviation, and shape deviation is designed for object detection. Finally, for the validity evaluation of the proposed method, experiments are performed on the insulator defect data set. It is shown in the results that the improved algorithm achieves a mask accuracy improvement of 5.82% and a detection speed of 37.4 FPS, which better complete the instance segmentation of insulator defect images.

Maize is one of the most important food crops in the field of agricultural development and food security in China. Among them, maize fruit and ear can directly determine the yield and quality of maize. Their phenotypic parameters (such as size and shape) can also be crucial indicators for the growth state of the plant. Fortunately, machine vision can be expected to serve as the maize phenotypic parameter acquisition and trait analysis, due to its objectivity, accuracy and speed, particularly with the application of artificial intelligence technology in agricultural production. Field inspection robots can be utilized to monitor the maize growth status in large-scale planting mode during this stage. This study aims to realize the high-throughput and automated acquisition of maize phenotypic parameters by unmanned inspection robots. A high-precision-speed balanced maize ear segmentation model, SwinT-YOLACT was proposed using YOLACT (you only look at coefficients) algorithm. Three optimization strategies were designed, according to the characteristics of the maize ear segmentation task. Firstly, SwinTransformer was used as the backbone feature extraction network of the improved model, where the self-attention mechanism of Transformer structure was integrated to enhance the global feature extraction capability; Secondly, a 3-layer effective channel attention mechanism was introduced before the feature pyramid network to eliminate the redundant feature information, in order to enhance the fusion of key features for the high accuracy of the improved model; Finally, the Mish activation function with better smoothing was used to replace the original Relu activation function, in order to further improve the segmentation accuracy at the original inference speed. In addition, the maize plant data was collected with the different environmental backgrounds and various maturity stages of maize fruit and maize ears in the field. Labelme annotation software was then adopted to manually label the data, according to the COCO dataset format. The number of samples was also expanded using data augmentation. A segmentation dataset of the maize fruit and ear was constructed for the model training on a deep learning network. The self-built segmentation dataset of maize ear was used to train and test the improved model. The experimental results show that the mask mean average precision was improved by 2.11 percentage points after introducing Swin-Transformer as the backbone feature extraction network, compared with the original YOLACT model. There was no influence on the segmentation speed. On this basis, the mask mean average precision was improved by 0.65 percentage points after introducing efficient channel attention before the feature pyramid network. The inference speed of the model was basically unchanged. The original model Relu activation function was replaced by Mish activation function, according to the first two experiments. The mask mean average precision was improved by 0.75 percentage points than before the replacement, whereas, the model inference speed was improved by 2.74 frames per second. SwinT-YOLACT was also used to compare with the YOLACT, YOLACT++, YOLACT-Edge, and Mask R-CNN segmentation models, all of which used the same experimental environment and training strategy. The verification results show that the mask mean average precision of SwinT-YOLACT reached 79.43%, which was 3.51, 3.38, and 7.88 percentage points higher than those of the original YOLACT, YOLACT++, and YOLACT-Edge, respectively, while only slightly lower than that of the Mask R-CNN model. The better performance of the improved models was then achieved in the segmentation task. In terms of segmentation speed, the inference speed of SwinTYOLACT was 35.44 frames per second, which was much better than that of Mask R-CNN model at 6.80 frames per second, and also improved by 3.39 and 2.58 frames per second, compared with the YOLACT and YOLACT++, respectively. In summary, SwinT-YOLACT can be expected for better segmentation of the maize fruit and ear in the unmanned inspection robot vision system. The finding can provide technical support for maize growth status monitoring. [ABSTRACT FROM AUTHOR]

When navigating ships in cold regions, sea ice concentration plays a crucial role in determining a ship's navigability. However, automatically extracting the sea ice concentration and floe size distribution remains challenging, due to the difficulty in detecting all the ice floes from the images captured in complex polar environments, particularly those that include both ships and sea ice. In this paper, we propose using the YOLACT network to address this issue. Cameras installed on the ship collect images during transit and an image dataset is constructed to train a model that can intelligently identify all the targets in the image and remove any noisy targets. To overcome the challenge of identifying seemingly connected ice floes, the non-maximum suppression (NMS) in YOLACT is improved. Binarization is then applied to process the detection results, with the aim of obtaining an accurate sea ice concentration. We present a color map and histogram of the associated floe size distribution based on the ice size. The speed of calculating the sea ice density of each image reaches 21 FPS and the results show that sea ice concentration and floe size distribution can be accurately measured. We provide a case study to demonstrate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

In order to overcome the problems of slow detection speed, low detection accuracy, dense wood stacks and easily obscured and overlooked, a segmentation method based on YOLACT_WOOD is proposed. YOLACT algorithm is proposed to explore the feasibility of a single-stage instance segmentation model for fast and accurate segmentation of whole-truck wood. In this study, based on the original YOLACT model, firstly, the ResNeXt network embedded with the CBAM attention mechanism module is used as the backbone network to imporve the feature extraction capability of the model; secondly, the image input size is increased to improve the detection ability of medium and small diameter class wood; then the CIoU bounding box regression loss function is used to improve the accuracy of bounding box regression; finally, DIoU is combined with Fast-NMS as a boundary box screening algorithm to improve the problem of false and missed detections. In this study, the YOLACT_WOOD algorithm is evaluated using five evaluation metrics: mAP, FPS, ${\mathrm{ IoU}}_{mask}$ , wood true detection rate, and parametric size, and the wood segmentation mask map is fitted and counted using the OpenCV library. The experimental results show that the mAP of this study method is improved by 5.6% compared to the original network, the ${\mathrm{ IoU}}_{mask}$ is improved by 2.6%, the FPS is improved by 14.7 frames/sec compared to the detection speed of the Mask R-CNN model, and the true detection rate of the logs in the test set reaches 96.61%, the false detection rate is 0.23%, and the parametric number of the model is not significantly improved. This result shows that the YOLACT_WOOD model not only ensures the segmentation speed but also improves the segmentation accuracy, solves the problem of false and omission, and the algorithm has strong robustness and generalisation ability.

The problem of small and multi-object polished rice image segmentation has always been one of importance and difficulty in the field of image segmentation. In the appearance quality detection of polished rice, image segmentation is a crucial part, directly affecting the results of follow-up physicochemical indicators. To avoid leak detection and inaccuracy in image segmentation qualifying polished rice, this paper proposes a new image segmentation method (YO-LACTS), combining YOLOv5 with YOLACT. We tested the YOLOv5-based object detection network, to extract Regions of Interest (RoI) from the whole image of the polished rice, in order to reduce the image complexity and maximize the target feature difference. We refined the segmentation of the RoI image by establishing the instance segmentation network YOLACT, and we eventually procured the outcome by merging the RoI. Compared to other algorithms based on polished rice datasets, this constructed method was shown to present the image segmentation, enabling researchers to evaluate polished rice satisfactorily. [ABSTRACT FROM AUTHOR]

Timely and reliable mapping of individual date palm trees is essential for their monitoring, health and risk assessment, pest control, and sustainable management of the date palm industry. This study presents an instance segmentation framework for large-scale detection and mapping of date palm trees using unmanned aerial vehicle (UAV)-based images. First, a data conversion framework is created to convert UAV image tiles and ground-truth vector data into annotation format of Common Objects in Context. Second, this study examines the efficacy of various instance segmentation models, namely, mask region convolutional neural network (Mask R-CNN), Mask Scoring R-CNN, You Only Look At CoefficientTs, Point-based Rendering, Segmenting Objects by Locations (SOLO), and SOLOv2) with varying residual learning networks (ResNets) in detecting and delineating individual date palm trees. Furthermore, the performance of two variants of Swin Transformer networks with a feature pyramid network (FPN) (Swin-small-FPN and Swin-tiny-FPN) as Mask RCNN network backbones was also evaluated. Third, we assess the generalizability of the evaluated instance segmentation models and backbones on different testing datasets with varying spatial resolutions. Results show that Mask R-CNN models based on Swin Transformers backbones outperform those with ResNets in the detection and segmentation of date palm trees with mAP50 of 92% and 91% and F-measures of 94% and 93%. Moreover, the Mask scoring R-CNN-based ResNet-50 and Mask R-CNN with a Swin-small-FPN backbone outperform the evaluated models and demonstrate great generalizability in different datasets with diverse spatial resolutions. The proposed instance segmentation framework provides an efficient tool for date palm tree mapping from multi-scale UAV-based images and is valuable and suitable for individual tree crown delineations and other earth-related applications. [ABSTRACT FROM AUTHOR]

Summary: Instance segmentation is a challenging task that requires both instance‐level and pixel‐level prediction and it has a wide range of applications in autonomous driving, video analysis, scene understandingand so on. The currently dominant instance segmentation methods have excellent accuracy, but they are slow, and the processing speed will be even less satisfactory if the input is a large‐scale image. In order to improve the efficiency and accuracy of instance segmentation of large‐scale images, this article modifies the backbone network based on YOLACT network, adds a multi‐information fusion module and provides an improved BiFPN method to achieve multi‐scale feature fusion, while adding two branches to the first level detector RetinaNet to achieve instance segmentation. The network model is tested on Cityscapes dataset and the results of the experiments show that the improved instance segmentation network in this article improves the accuracy while ensuring the speed of segmentation. The optimized network model size was reduced by 17% compared to YOLACT, and the mAP, mAP50, and mAP75 were improved by 18.3%, 32.1%, and 24.6%, respectively. [ABSTRACT FROM AUTHOR]

After years of service, bridges could lose their expected functions. Considering the significant number of bridges and the adverse inspecting environment, the urgent requirement for timely and efficient inspection solutions, such as computer vision techniques, have been attractive in recent years, especially for those bridge components with poor accessibility. In this paper, a lightweight procedure for bridge apparent-defect detection is proposed, including a crack annotation method and crack detection. First of all, in order to save computational costs and improve generalization performance, we propose herein a relative-breakpoint annotation method to build a crack instance segmentation dataset, a critical process for a supervised vision-based crack detection method. Then, the trained models based on classic Mask RCNN and Yolact are transferred to evaluate the effectiveness of the proposed method. To verify the correctness, universality and generality of the proposed crack-detection framework, approximately 800 images are used for model training, while nearly 100 images are saved for validation. Results show that the crack instance segmentation model can achieve a level of 90% for both accuracy and recall values, with a limited dataset.

In intelligent driving, the detection and ranging of drivable area are key technologies in path planning. In order to realize quick and accurate detection and segmentation of drivable area, we adopt YOLACT_ResNet38_TFPN network as an improved design of YOLACT. The original YOLACT has ResNet101 residual structure. We reduce the network layers to ResNet38 and add C6 and C7 with larger receiving field to improve the detection speed in the drivable area. Moreover, the FPN (feature pyramid network) is designed as a structure of three sizes to match C6 and C7. According to the characteristics of road image, the aspect ratio of three anchor points is redesigned to further improve the detection accuracy and speed. A univariate linear regression model is designed to accurately calculate the distance of the drivable area. The model parameters are achieved by multivariate linear fitting method based on multiple sets of distance measurement data. Finally, YOLACT_ResNet38_TFPN's FPS for the drivable area is 46.13, the box mAP is 62.36, and the mask mAP is 61.36. The maximum ranging error of this method within 96.90 m is 5.3 m. It can quickly and accurately measure the drivable distance and provide reasonable path driving suggestions for intelligent driving. [ABSTRACT FROM AUTHOR]

Thousands of tonnes of litter enter the ocean every day, posing a significant threat to marine life and ecosystems. While floating and beach litter are often in the spotlight, about 70% of marine litter eventually sinks to the seafloor, making underwater litter the largest accumulation of marine litter that often goes undetected. Plastic debris makes up the majority of ocean litter and is a known source of microplastics in the ocean. This paper focuses on the detection of ocean plastic using neural network models. Two neural network models will be trained, i.e., YOLACT and the Mask R-CNN, for the instance segmentation of underwater litter in images. The models are trained on the TrashCAN dataset, using pre-trained model weights trained using COCO. The trained neural network could achieve a mean average precision (mAP) of 0.377 and 0.365 for the Mask R-CNN and YOLACT, respectively. The lightweight nature of YOLACT allows it to detect images at up to six times the speed of the Mask R-CNN, while only making a comparatively smaller trade-off in terms of performance. This allows for two separate applications: YOLACT for the collection of litter using autonomous underwater vehicles (AUVs) and the Mask R-CNN for surveying litter distribution.

When navigating ships in cold regions, sea ice concentration plays a crucial role in determining a ship’s navigability. However, automatically extracting the sea ice concentration and floe size distribution remains challenging, due to the difficulty in detecting all the ice floes from the images captured in complex polar environments, particularly those that include both ships and sea ice. In this paper, we propose using the YOLACT network to address this issue. Cameras installed on the ship collect images during transit and an image dataset is constructed to train a model that can intelligently identify all the targets in the image and remove any noisy targets. To overcome the challenge of identifying seemingly connected ice floes, the non-maximum suppression (NMS) in YOLACT is improved. Binarization is then applied to process the detection results, with the aim of obtaining an accurate sea ice concentration. We present a color map and histogram of the associated floe size distribution based on the ice size. The speed of calculating the sea ice density of each image reaches 21 FPS and the results show that sea ice concentration and floe size distribution can be accurately measured. We provide a case study to demonstrate the effectiveness of the proposed approach.

The navel orange planting area in Gannan region of China has reached the first in the world, with an annual output of one million tons. However, most navel oranges were picked manually with high labor intensity and low efficiency. It was very urgent to introduce intelligent picking robots. In the process of navel orange picking, the picking robot needed to obtain the spatial position and fruit size information of navel orange fruit in real time to achieve efficient and rapid intelligent grading picking. This study proposed an algorithm framework--OrangePointSge based on the improved YOLACT model and the least-square sphere fitting method. The real-time instance segmentation algorithm YOLACT generated an instance mask and cropped the registered depth image to obtain the fruit depth point cloud. Then, the least-square method was used to fit the navel orange fruit shape, and the centroid coordinates and radius under the camera coordinate system were obtained. Firstly, the RGB-D data of navel orange fruit was collected by Microsoft's latest consumer-level depth camera (Azure Kinect DK), and the navel orange fruit instance segmentation dataset and enhancement dataset were established and free to open source. The dataset contained 2 178 images, including 8 712 samples. Among those samples, the number of non-occluded fruit was 4 682, and the number of slightly occluded fruit was 4 030. Then the YOLACT algorithm was improved to adapt to the detection of navel orange fruit, and the instance mask of the fruit was output. We modified the original ResNet (deep residual network) backbone network to HRNet (High-Resolution Net) to simplify the model, and used the hrnet_w48 structure for feature extraction to improve the model detection accuracy. We also optimized the non-maximum suppression process to improve the detection speed. At the same time, several groups of comparative tests were also set up, including the performance of YOLACT with different backbone networks and the performance of different algorithms in different scenarios. To improve the generalization ability of the training model, we carried out a series of enhancement algorithms for color images, mainly including scale change, color change, brightness change, and adding noise. For long-range small targets, we used oversampling method for enhancement training to improve their detection accuracy. Therefore, the recognition speed and mask accuracy of navel orange were higher than other algorithms. The detection speed was 44.63 frames/s and the average accuracy was 31.15%. The mask generated by improved YOLACT will be cropped with the depth image registered with the color image to obtain the depth information of each fruit to generate the fruit depth point cloud from the depth image. The least-square method was used to fit the point cloud to obtain the centroid coordinates and radius of the fruit, which were used to guide the robot to carry out grading picking. In addition, we added the comparison with RANSAC (Random Sample Consensus) algorithm. The results show that the least-square method is faster than RANSAC while ensuring the appropriate accuracy. When the number of point clouds was 1 400-2 000, the fitting time was 1.99 ms, the positioning error was 0.49 cm, the fitting radius root mean square error was 0.43 cm, and the volume root mean square error was 52.6 mL. Through the experiment of setting different distances on the fitting accuracy, it was found that when the number of point clouds was more than 800 and the distance was less than 1.0 m, the positioning error was controlled within 0.46 cm. Finally, by introducing parallel computing and combining the above two processing flows, the overall processing speed of OrangePointSge was 29.4 frames/s, which can better balance the accuracy and speed. Thus, the proposed algorithm is conducive to practical application and engineering deployment. [ABSTRACT FROM AUTHOR]

As reconstruction and redevelopment accelerate, the generation of construction waste increases, and construction waste treatment technology is being developed accordingly, especially using artificial intelligence (AI). The majority of AI research projects fail as a consequence of poor learning data as opposed to the structure of the AI model. If data pre-processing and labeling, i.e., the processes prior to the training step, are not carried out with development purposes in mind, the desired AI model cannot be obtained. Therefore, in this study, the performance differences of the construction waste recognition model, after data pre-processing and labeling by individuals with different degrees of expertise, were analyzed with the goal of distinguishing construction waste accurately and increasing the recycling rate. According to the experimental results, it was shown that the mean average precision (mAP) of the AI model that trained on the dataset labeled by non-professionals was superior to that labeled by professionals, being 21.75 higher in the box and 26.47 in the mask, on average. This was because it was labeled using a similar method as the Microsoft Common Objects in Context (MS COCO) datasets used for You Only Look at Coefficients (YOLACT), despite them possessing different traits for construction waste. Construction waste is differentiated by texture and color; thus, we augmented the dataset by adding noise (texture) and changing the color to consider these traits. This resulted in a meaningful accuracy being achieved in 25 epochs—two fewer than the unreinforced dataset. In order to develop an AI model that recognizes construction waste, which is an atypical object, it is necessary to develop an explainable AI model, such as a reconstruction AI network, using the model's feature map or by creating a dataset with weights added to the texture and color of the construction waste. [ABSTRACT FROM AUTHOR]

The problem of small and multi-object polished rice image segmentation has always been one of importance and difficulty in the field of image segmentation. In the appearance quality detection of polished rice, image segmentation is a crucial part, directly affecting the results of follow-up physicochemical indicators. To avoid leak detection and inaccuracy in image segmentation qualifying polished rice, this paper proposes a new image segmentation method (YO-LACTS), combining YOLOv5 with YOLACT. We tested the YOLOv5-based object detection network, to extract Regions of Interest (RoI) from the whole image of the polished rice, in order to reduce the image complexity and maximize the target feature difference. We refined the segmentation of the RoI image by establishing the instance segmentation network YOLACT, and we eventually procured the outcome by merging the RoI. Compared to other algorithms based on polished rice datasets, this constructed method was shown to present the image segmentation, enabling researchers to evaluate polished rice satisfactorily.

As reconstruction and redevelopment accelerate, the generation of construction waste increases, and construction waste treatment technology is being developed accordingly, especially using artificial intelligence (AI). The majority of AI research projects fail as a consequence of poor learning data as opposed to the structure of the AI model. If data pre-processing and labeling, i.e., the processes prior to the training step, are not carried out with development purposes in mind, the desired AI model cannot be obtained. Therefore, in this study, the performance differences of the construction waste recognition model, after data pre-processing and labeling by individuals with different degrees of expertise, were analyzed with the goal of distinguishing construction waste accurately and increasing the recycling rate. According to the experimental results, it was shown that the mean average precision (mAP) of the AI model that trained on the dataset labeled by non-professionals was superior to that labeled by professionals, being 21.75 higher in the box and 26.47 in the mask, on average. This was because it was labeled using a similar method as the Microsoft Common Objects in Context (MS COCO) datasets used for You Only Look at Coefficients (YOLACT), despite them possessing different traits for construction waste. Construction waste is differentiated by texture and color; thus, we augmented the dataset by adding noise (texture) and changing the color to consider these traits. This resulted in a meaningful accuracy being achieved in 25 epochs—two fewer than the unreinforced dataset. In order to develop an AI model that recognizes construction waste, which is an atypical object, it is necessary to develop an explainable AI model, such as a reconstruction AI network, using the model’s feature map or by creating a dataset with weights added to the texture and color of the construction waste.

The demand for categorising technology that requires minimum manpower and equipment is increasing because a large amount of waste is produced during the demolition and remodelling of a structure. Considering the latest trend, applying an artificial intelligence (AI) model for automatic categorisation is the most efficient method. However, it is difficult to apply this technology because research has only focused on general domestic waste. Thus, in this study, we delineate the process for developing an AI model that differentiates between various types of construction waste. Particularly, solutions for solving difficulties in collecting learning data, which is common in AI research in special fields, were also considered. To quantitatively increase the amount of learning data, the Fréchet Inception Distance method was used to increase the amount of learning data by two to three times through augmentation to an appropriate level, thus checking the improvement in the performance of the AI model.

This study develops an efficient approach for precise channel frame detection in complex backgrounds, addressing the critical need for accurate drone navigation. Leveraging YOLACT and group regression, our method outperforms conventional techniques that rely solely on color information. We conducted extensive experiments involving channel frames placed at various angles and within intricate backgrounds, training the algorithm to effectively recognize them. The process involves initial edge image detection, noise reduction through binarization and erosion, segmentation of channel frame line segments using the Hough Transform algorithm, and subsequent classification via the K-means algorithm. Ultimately, we obtain the regression line segment through linear regression, enabling precise positioning by identifying intersection points. Experimental validations validate the robustness of our approach across diverse angles and challenging backgrounds, making significant advancements in UAV applications., Competing Interests: The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Yu-Shiuan Tsai reports financial support was provided by Ministry of Science and Technology (10.13039/501100004663MOST), Taiwan. Yu-Shiuan Tsai reports financial support was provided by National Science and Technology Council (NSTC), Taiwan., (© 2024 The Authors.)

Detection and tracking of all major parts of pig body could be more productive to help to analyze pig behavior. To achieve this goal, a real-time algorithm based on You Only Look At CoefficienTs (YOLACT) was proposed. A pig body was divided into ten parts: one head, one trunk, four thighs and four shanks. And the key points of each part were calculated by the novel algorithm, which was based mainly on combination of the Zhang-Suen thinning algorithm and Gravity algorithm. The experiment results showed that these parts of pig body could be detected and tracked, and their contributions to overall pig activity could also be sought out. The detect accuracy of the algorithm in the data set could reach up to 90%, and the processing speed to 30.5 fps. Furthermore, the algorithm was robust and adaptive. [ABSTRACT FROM AUTHOR]

The shot-type decision is a very important pre-task in movie analysis due to the vast information, such as the emotion, psychology of the characters, and space information, from the shot type chosen. In order to analyze a variety of movies, a technique that automatically classifies shot types is required. Previous shot type classification studies have classified shot types by the proportion of the face on-screen or using a convolutional neural network (CNN). Studies that have classified shot types by the proportion of the face on-screen have not classified the shot if a person is not on the screen. A CNN classifies shot types even in the absence of a person on the screen, but there are certain shots that cannot be classified because instead of semantically analyzing the image, the method classifies them only by the characteristics and patterns of the image. Therefore, additional information is needed to access the image semantically, which can be done through semantic segmentation. Consequently, in the present study, the performance of shot type classification was improved by preprocessing the semantic segmentation of the frame extracted from the movie. Semantic segmentation approaches the images semantically and distinguishes the boundary relationships among objects. The representative technologies of semantic segmentation include Mask R-CNN and Yolact. A study was conducted to compare and evaluate performance using these as pretreatments for shot type classification. As a result, the average accuracy of shot type classification using a frame preprocessed with semantic segmentation increased by 1.9%, from 93% to 94.9%, when compared with shot type classification using the frame without such preprocessing. In particular, when using ResNet-50 and Yolact, the classification of shot type showed a 3% performance improvement (to 96% accuracy from 93%). [ABSTRACT FROM AUTHOR]

Conventionally, the number of steel rebars at construction sites is manually counted by workers. However, this practice gives rise to several problems: it is slow, human-resource-intensive, time-consuming, error-prone, and not very accurate. Consequently, a new method of quickly and accurately counting steel rebars with a minimal number of workers needs to be developed to enhance work efficiency and reduce labor costs at construction sites. In this study, the authors developed an automated system to estimate the size and count the number of steel rebars in bale packing using computer vision techniques based on a convolutional neural network (CNN). A dataset containing 622 images of rebars with a total of 186,522 rebar cross sections and 409 poly tags was established for segmentation rebars and poly tags in images. The images were collected in a full HD resolution of 1920 × 1080 pixels and then center-cropped to 512 × 512 pixels. Moreover, data augmentation was carried out to create 4668 images for the training dataset. Based on the training dataset, YOLACT-based steel bar size estimation and a counting model with a Box and Mask of over 30 mAP was generated to satisfy the aim of this study. The proposed method, which is a CNN model combined with homography, can estimate the size and count the number of steel rebars in an image quickly and accurately, and the developed method can be applied to real construction sites to efficiently manage the stock of steel rebars.

• Multimodal hierarchical fusion method based on attention mechanism is proposed. • Improved recognition performance of tomato stem against similar-colored background. • Proposed method improves recognition by fusing RGB and NIR image features. • Multimodal loss function reduces the influence of information mutual exclusion. • Local depth separable convolution and Conv-BN layer merging reduce inference time. The tomato plant's main-stem is a feasible lead for robotic searching the grows discretely-growing targets of harvesting, pruning or pollinating. Owing to the highlighted reflection characteristics of the main-stem in the near-infrared (NIR) waveband, this study proposes a multimodal hierarchical fusion method (YOLACTFusion) based on the attention mechanism, to achieve an instance segmentation of the main-stem from similar-colored differentiation (i.e., green leaf and green fruit) in robotic vision systems. The model inputs RGB images and 900–1100 nm NIR images into two ResNet50 backbone networks and uses a parallel attention mechanism to fuse feature maps of various scales together into the head network, to improve the segmentation performance of the main-stem of RGB images. The loss function for the multimodal image weights the original loss on the RGB image and the position offset loss and classification loss on the NIR image. Furthermore, the local depthwise separable convolution is used for the backbone network, and Conv-BN layers are merged to reduce the computational complexity. The results show that the precision and recall of YOLACTFusion of the main-stem detection, respectively reached 93.90 % and 62.60 %; and the precision and recall of instance segmentation reached 95.12 % and 63.41 %, respectively. Compared to YOLACT, the mean average precision (mAP) of YOLACTFusion is increased from 39.20 % to 46.29 %, the model size is reduced from 199.03 MB to 165.52 MB , while the image processing efficiency remains similar. The overall results show that the multimodal instance segmentation method proposed in this study significantly improves the detection and segmentation of tomato main-stems under a similar-colored background, which would be a potential method for improving agricultural robot's visual perception. [ABSTRACT FROM AUTHOR]

• UAV RGB videos were employed to extract grape phenotypic traits at an early stage of development. • It detects and tracks white grape bunches in a vineyard with leaf occlusion and challenging sunny conditions. • It detects grape berries within the detected bunches. • Early extraction of phenotyping traits may support vineyard managers to monitor the sanitary status of the vine and get information on the yield to be harvested. Grapevine phenotyping is the process of determining the physical properties (e.g., size, shape, and number) of grape bunches and berries. Grapevine phenotyping information provides valuable characteristics to monitor the sanitary status of the vine. Knowing the number and dimensions of bunches and berries at an early stage of development provides relevant information to the winegrowers about the yield to be harvested. However, the process of counting and measuring is usually done manually, which is laborious and time-consuming. Previous studies have attempted to implement bunch detection on red bunches in vineyards with leaf removal and surveys have been done using ground vehicles and handled cameras. However, Unmanned Aerial Vehicles (UAV) mounted with RGB cameras, along with computer vision techniques offer a cheap, robust, and timesaving alternative. Therefore, Multi-object tracking and segmentation (MOTS) is utilized in this study to determine the traits of individual white grape bunches and berries from RGB videos obtained from a UAV acquired over a commercial vineyard with a high density of leaves. To achieve this goal two datasets with labelled images and phenotyping measurements were created and made available in a public repository. PointTrack algorithm was used for detecting and tracking the grape bunches, and two instance segmentation algorithms - YOLACT and Spatial Embeddings - have been compared for finding the most suitable approach to detect berries. It was found that the detection performs adequately for cluster detection with a MODSA of 93.85. For tracking, the results were not sufficient when trained with 679 frames.This study provides an automated pipeline for the extraction of several grape phenotyping traits described by the International Organization of Vine and Wine (OIV) descriptors. The selected OIV descriptors are the bunch length, width, and shape (codes 202, 203, and 208, respectively) and the berry length, width, and shape (codes 220, 221, and 223, respectively). Lastly, the comparison regarding the number of detected berries per bunch indicated that Spatial Embeddings assessed berry counting more accurately (79.5%) than YOLACT (44.6%). [ABSTRACT FROM AUTHOR]

The shot-type decision is a very important pre-task in movie analysis due to the vast information, such as the emotion, psychology of the characters, and space information, from the shot type chosen. In order to analyze a variety of movies, a technique that automatically classifies shot types is required. Previous shot type classification studies have classified shot types by the proportion of the face on-screen or using a convolutional neural network (CNN). Studies that have classified shot types by the proportion of the face on-screen have not classified the shot if a person is not on the screen. A CNN classifies shot types even in the absence of a person on the screen, but there are certain shots that cannot be classified because instead of semantically analyzing the image, the method classifies them only by the characteristics and patterns of the image. Therefore, additional information is needed to access the image semantically, which can be done through semantic segmentation. Consequently, in the present study, the performance of shot type classification was improved by preprocessing the semantic segmentation of the frame extracted from the movie. Semantic segmentation approaches the images semantically and distinguishes the boundary relationships among objects. The representative technologies of semantic segmentation include Mask R-CNN and Yolact. A study was conducted to compare and evaluate performance using these as pretreatments for shot type classification. As a result, the average accuracy of shot type classification using a frame preprocessed with semantic segmentation increased by 1.9%, from 93% to 94.9%, when compared with shot type classification using the frame without such preprocessing. In particular, when using ResNet-50 and Yolact, the classification of shot type showed a 3% performance improvement (to 96% accuracy from 93%).

Accurate and rapid perception of fruit bunch posture is necessary for the cherry tomato harvesting robot to successfully achieve the bunch's holding and separating. According to the postural relationship of the fruit bunch, bunch pedicel, and plant' main-stem, the robotic end-effector's holding region and approach path could be determined, which were important for successful picking operation. The main goal of this research was to propose a multitask-aware network (MTA-YOLACT), which simultaneously performed region detection on fruit bunch, and region segmentation on pedicel and main-stem. The MTA-YOLACT extended from the pre-trained YOLACT model, included two detector branch networks for detection and instance segment, which shared the same backbone network, and the loss function with weighting coefficients of the two branches was adopted to balance the multi-task learning, according to multi-task's homoscedastic uncertainty during the model training. Furthermore, in order to cluster the fruit bunch, pedicel and main-stem from the same bunch target, a classification and regression tree (CART) model was built, based on the region's positional relationship from the MTA-YOLACT output. An image dataset of cherry tomato plants in China greenhouse was built to training and test the model. The results indicated a promising performance of the proposed network, with an F1-score of 95.4% on detecting fruit bunches and the mean Average Precision of 38.7% and 51.9% on the instance segmentation of pedicel and main-stem, which was 1.1% and 3.5% more than original YOLACT. Beyond that, our approach performed a real-time detection and instance segmentation of 13.3 frames per second (FPS). The whole bunch could be identified by the CART model with an average accuracy of 99.83% and the time cost of 9.53 ms. These results demonstrated the research could be a viable support to the harvesting robot's vision unit development and the end-effector's motion planning in the future research. • A Multi-task Awareness Network (MTA-YOLACT) improved fruit bunch identification. • MTA-YOLACT detected and segmented regions on the fruit bunch at the same time. • Analysis of features satisfying subordination constraints in multi-fruit bunches. • A decision model clustered the fruit bunch, pedicel and main-stem from same bunch. • Experiments show maintain excellent at the real-time and accuracy. [ABSTRACT FROM AUTHOR]

The condition monitoring of railway track line is one of the fundamental tasks to ensure the safety of the railway transportation system. Railway track line is mainly made up of tracks, fasteners, bolts, backing plates, and so on. Given the requirements for rapid and accurate inspection, an innovative and intelligent method for multi-component identification and common defect detection of railway track line is investigated based on instance segmentation. More specifically, a railway track line image (RTL-I) dataset is constructed and annotated manually in this paper. After that, based on the work of YOLACT and YOLACT++, combined with prior knowledge, a railway track line image segmentation model (RTLSeg for short) is proposed. Firstly, taking the characteristics of the objects in the RTL-I dataset, preset anchors are redesigned and a feature enhanced module is introduced in the prediction head to improve the detection and segmentation accuracy of the model. Secondly, to strengthen the internal information propagation within the model, PaFPN (path aggregation feature pyramid network) is applied instead of FPN in RTLSeg. Thirdly, with the help of CoordConv, Coord-Protonet is presented to add position awareness explicitly to the model for more robust and higher quality prototype masks. Finally, to further improve the model performance, the attention mechanism is explored and a novel spatial attention-guided bounding box branch is employed in the enhanced prediction head. Both quantitative and qualitative experimental results show that the proposed method is feasible in detecting and segmenting multi-component and common defects of railway track line, and outperforms the compared baseline models. In particular, RTLSeg is able to achieve 91.35 bbox mAP and 91.60 mask mAP with the customized dataset. Meanwhile, the average inference speed reaches 13.07 fps. The average detection accuracy and recall are 100% and 99.83%, respectively. Furthermore, the effectiveness of each optimized part of the proposed RTLSeg model is demonstrated by additional ablation study. • Pixel-level multi-component inspection of railway track line based on image instance segmentation is investigated for the first time. • Based on YOLACT and YOLACT++, a railway track line image segmentation model (RTLSeg for short) is proposed, which achieves better detection and segmentation performance than the compared baseline models, and is more suitable for the scenario of the railway track line. • The proposed RTLSeg model enables accurate and fast detection of railway track line components and common defects at the pixel level. [ABSTRACT FROM AUTHOR]

The identification and digitization of floating ice are of great significance to developing numerical ice load model of ships and marine structures in managed ice fields. The circumferential crack method is a commonly-used method to simulate the level ice breaking process for a polar ship. The crack size, including radius and open-angle, is an import parameter that should be considered in the analysis of ice-structure interactions. It is usually determined by the empirical formulas, which may introduce many uncertainties in the selection of input parameters. In this paper, we adopt a deep learning YOLACT model to identify broken ice blocks from an image, where an image processing algorithm is newly developed to estimate circumferential crack size for each identified ice blocks. The numerical simulation shows that the accuracy of the radius and open-angle estimated by the present method can reach up to 97.37% and 96.66%, respectively. The present tool could be used to assist design and operations for marine structures in ice-infested waters. • We propose a new image processing algorithm to estimate circumferential crack size for each identified ice blocks. • The crack size, including radius and open-angle, can be accurately and automatically measured. • The uncertainties introduced in the empirical formulas are improved by the present method. • The different fitting result of least square method and the proposed method is analyzed and the accuracy is validated by the Hough circle detection. • We quantitatively analyse the difference of actual ice-breaking radius and fitting ice-breaking radius. [ABSTRACT FROM AUTHOR]