Search

The Train Rolling-Stock Examination (TRSE) is a safety examination process that physically examines the bogie parts of a moving train, typically at speeds over 30 km/h. Currently, this inspection process is done manually by railway personnel in many countries to ensure safety and prevent interruptions to rail services. Although many earlier attempts have been made to semi-automate this process through computer-vision models, these models are iterative and still require manual intervention. Consequently, these attempts were unsuitable for real-time implementations. In this work, we propose a detection model by utilizing a deep-learning based classifier that can precisely identify bogie parts in real-time without manual intervention, allowing an increase in the deployability of these inspection systems. We implemented the Anchor-Free Yolov8 (AFYv8) model, which has a decoupled-head module for recognizing bogie parts. Additionally, we incorporated bogie parts tracking with the AFYv8 model to gather information about any missing parts. To test the effectiveness of the AFYv8-model, the bogie videos were captured at three different timestamps and the result shows the increase in the recognition accuracy of TRSE by 10 % compared to the previously developed classifiers. This research has the potential to enhance railway safety and minimize operational interruptions. [ABSTRACT FROM AUTHOR]

In today's fast-moving digital era, video technology plays an important role. An effective video summarising approach is urgently needed to handle a lot of video data due to the ever-growing number of video content. In this paper, the authors have proposed, A hybrid summarization methodology for video summary evaluation using multimedia features (Text, Images, and Audio) that assess how well a video summary can keep the ranking of vital video frames, semantic data, and audio present in the original video. Video summary can be evaluated by ranking text, audio, and semantics of video frames, giving more accurate summarisation results. The proposed methodology works in three phases: The first part takes the text in the video, the second phase takes the audio to the file, and the last phase focuses on the video frames rather than images in the video. TVSum dataset has been used for the experimentation. F1 has been used as the evaluation metric for checking the efficacy and efficiency of the proposed methodology. The result shows that the proposed hybrid model achieves the highest F1 score of 69.9% and saves 75-80% of user time in watching video summaries instead of the whole video. [ABSTRACT FROM AUTHOR]

Copyright of Arquivos Brasileiros de Oftalmologia is the property of Arquivos Brasileiros de Oftalmologia 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.)

Objective: In this work, we assess the potential of computer vision techniques for age estimation of Gulf Menhaden Brevoortia patronus scales. Scales are the primary structure used for the age determination of Gulf Menhaden, and the ageing process can be labor intensive. Gulf Menhaden is the second‐largest fishery by weight in the United States, with average annual landings from 2018 to 2022 of 449,540 metric tons, and is assessed with age‐structured models that require information about the age structure of the catch and the stock. Methods: We used convolutional neural networks and deep neural networks to classify the age from images of Gulf Menhaden scales from three different sets of images of scales. The first set of data consists of images of scales from fish at ages 0 and 1 year. The second set of data consists of images of scales from fish at ages 0–4 years. The last set of data consists of images of scales from fish of ages 0, 1, and 2 years and includes only images of scales for which there is agreement by readers of age estimates derived from analyzing sagittal otoliths and scales from the same individual. Result: The classification of ages was best when using a convolutional neural network model on the first data set. The poorest classification was for the model using deep neural networks with the second data set. Conclusion: Although we show that computer vision has promise for age determination from fish scale samples, our results indicate that considerable work must be done for wide adoption of the approach. With the continuous enhancements of computer vision models, improvements in the quality of scale images, and the accumulation of larger sets of scale images that can be used to train machine learning models, we believe that using computer vision can serve to reduce processing time and increase the accuracy of age estimates. Impact statementComputer vision techniques have been widely employed for image analysis but remain novel for fish age determination in a management context. In this work, we show that the approach has promise and document some of the hurdles that must be overcome prior to widespread use. [ABSTRACT FROM AUTHOR]

Location-based social media data provides a new perspective for understanding the relationship between human behavior and urban environments. However, further research is needed to determine the application of cognitive computing in urban environments and physical activities. This study proposes a cognitive computing framework for urban environments and human activities that extracts knowledge from structured and unstructured data through natural language processing and computer vision techniques. This paper utilizes a Naive Bayes Model constructed based on random reviews, as well as semantic segmentation and instant segmentation algorithms based on convolutional neural networks to obtain information about urban environments and human behavior from social media data and other geospatial resources. This study examines the relationship between the urban environment and residents' activity, including spatiotemporal behavior, public sentiment, and physical activity. The study found statistically significant results in subgroup analyses regarding the effects of urban environments on sentiment and physical activity, which also exhibited a strong social gradient consistent with traditional findings. This study validates the feasibility of using cognitive computing based on social media data to explore environmental behaviors, providing technical support for updating health promotion policies. [ABSTRACT FROM AUTHOR]

The connecting buckles of underground transportation belts are a critical component. These components are susceptible to damage from material impact and pulley wear-induced damage, which can compromise industrial safety and transportation efficiency. While the importance of detecting defects in connecting buckles is self-evident, effective methods to solve this problem are currently lacking. To fill this gap, this study proposed a defect detection method using an enhanced version of the YOLOv7 model. The process incorporated a target detection head, an attention mechanism, and a loss function to enhance the model's generalization ability and robustness under complex working environments and diverse defect types. Through experimental validation using a self-constructed U38 buckle dataset, the method achieved an 83.2% detection rate for various defect types, with critical defects identified with an accuracy of up to 98.1%. The enhanced YOLOv7_m model exhibited strong performance in defect detection, providing valuable technical support for improving the safety and efficiency of coal mine transportation. This research provides a valuable reference for further optimizing the diagnosis of connection buckle defects based on deep learning. [ABSTRACT FROM AUTHOR]

To address the accuracy limitations of existing safety helmet detection algorithms in complex environments, we propose an enhanced YOLOv8 algorithm, called YOLOv8- CSS. We introduce a Coordinate Attention (CA) mechanism in the backbone network to improve focus on safety helmet regions in complex backgrounds, suppress irrelevant feature interference, and enhance detection accuracy. We also incorporate the SEAM module to improve the detection and recognition of occluded objects, increasing robustness and accuracy. Additionally, we design a fine-neck structure to fuse features of different sizes from the backbone network, reducing model complexity while maintaining detection accuracy. Finally, we adopt the Wise-IoU loss function to optimize the training process, further enhancing detection accuracy. Experimental results show that YOLOv8-CSS significantly improves detection performance in general scenarios, complex backgrounds, and for distant small objects. YOLOv8-CSS improves precision, recall, mAP@0.5, and mAP@0.5:0.95 by 1.67%, 5.55%, 3.38%, and 5.87%, respectively, compared to YOLOv8n. Our algorithm also reduces model parameters by 21.25% and computational load by 15.89%. Comparisons with other mainstream object detection algorithms validate our approach's effectiveness and superiority. [ABSTRACT FROM AUTHOR]

To demonstrate the use of surgical intelligence to routinely and automatically assess the proportion of time spent outside of the patient's body (out-of-body—OOB) in laparoscopic gynecological procedures, as a potential basis for clinical and efficiency-related insights. A retrospective analysis of videos of laparoscopic gynecological procedures. Two operating rooms at the Gynecology Department of a tertiary medical center. All patients who underwent laparoscopic gynecological procedures between January 1, 2021 and December 31, 2022 in those two rooms. A surgical intelligence platform installed in the two rooms routinely captured and analyzed surgical video, using AI to identify and document procedure duration and the amount and percentage of time that the laparoscope was withdrawn from the patient's body per procedure. A total of 634 surgical videos were included in the final dataset. The cumulative time for all procedures was 639 hours, of which 48 hours (7.5%) were OOB segments. Average OOB percentage was 8.7% (SD = 8.7%) for all the procedures and differed significantly between procedure types (p <.001), with unilateral and bilateral salpingo-oophorectomies showing the highest percentages at 15.6% (SD = 13.3%) and 13.3% (SD = 11.3%), respectively. Hysterectomy and myomectomy, which do not require the endoscope to be removed for specimen extraction, showed a lower percentage (mean = 4.2%, SD = 5.2%) than the other procedures (mean = 11.1%, SD = 9.3%; p <.001). Percentages were lower when the operating team included a senior surgeon (mean = 8.4%, standard deviation = 9.2%) than when it did not (mean = 10.1%, standard deviation = 6.9%; p <.001). Surgical intelligence revealed a substantial percentage of OOB segments in laparoscopic gynecological procedures, alongside associations with surgeon seniority and procedure type. Further research is needed to evaluate how laparoscope removal affects postoperative outcomes and operational efficiency in surgery. [ABSTRACT FROM AUTHOR]

Underwater concrete structure crack detection and structural health condition assessment based on image processing is a challenging task. The complex underwater environment and severe image degradation seriously affect the accuracy of crack detection. To solve these problems, a monocular vision and image-enhanced fractal-based fractal science based on computer vision and image processing techniques are proposed to carry out a non-contact detection study of underwater concrete cracks. In this study, a four-level structural health condition was established to assist in underwater crack measurement and safety assessment. The box-counting method was used as a practical tool to calculate the fractal dimension. To verify the effective distance of the algorithm, three distances of 0.5 m,0.8 m, and 1.2 m were set. The results show that the method proposed in this study can effectively detect cracks in submerged concrete members within 0.6 m and help managers correctly determine the health of the structure using the fractal dimension. [ABSTRACT FROM AUTHOR]

Recent years have seen an explosion in the application of artificial intelligence and machine learning (AI/ML) to practical problems from computer vision to game playing to algorithm design. This growth has been mirrored and, in many ways, been enabled by the development and maturity of publicly-available software packages that make model building, training, and testing easier than ever. While these packages provide tremendous power and flexibility to users, and greatly facilitate learning and deploying AI/ML techniques, they and the models they provide are extremely complicated and as a result can present a number of subtle but serious pitfalls. This paper presents three examples where obscure settings or bugs in these packages dramatically changed model behavior or performance – one from a deep learning regression application, one from reinforcement learning, and one from computer vision classification. These examples illustrate the importance of thinking carefully about the results that a model is producing and carefully checking each step in its development before trusting its output. [ABSTRACT FROM AUTHOR]

This work introduces a revolutionary YOLOv9 deep learning architecture-based method for facial emotion identification. Happy, Sad, Angry, Disgust, Natural, and Surprise are the six emotional classifications the research divides the 21,263 images into which the research divides. The model was trained on 88% of the dataset; 4% was used for testing and 8% for validation. The preprocessing operations were auto-orientation, zoom, rotation, and 416x416 pixel scaling. Training time was cut down significantly because the experiment was carried out with a T4 GPU from Kaggle. Five epochs later, the average mAP score of 0.85, average precision of 0.74, and average recall of 0.84 indicate encouraging emotional performance. At a mAP score of 0.98, the model showed exceptionally high accuracy in identifying disgust. With its reliable and effective approach to real-time emotion recognition in various applications, this work advances the field of affective computing. [ABSTRACT FROM AUTHOR]

The integration of artificial intelligence technology with medicine has rapidly evolved, with increasing demands for quality of life. However, falls remain a significant risk leading to severe injuries and fatalities, especially among the elderly. Therefore, the development and application of computer vision-based fall detection technologies have become increasingly important. In this paper, firstly, the keypoint detection algorithm ViTPose++ is used to obtain the coordinates of human body keypoints from the camera images. Human skeletal feature maps are generated from this keypoint coordinate information. Meanwhile, human dense feature maps are produced based on the DensePose algorithm. Then, these two types of feature maps are confused as dual-channel inputs for the model. The convolutional gated recurrent unit is introduced to extract the frame-to-frame relevance in the process of falling. To further integrate features across three dimensions (spatio-temporal-channel), a dual-channel fall detection algorithm based on video streams is proposed by combining the Convolutional Block Attention Module (CBAM) with the ConvGRU. Finally, experiments on the public UR Fall Detection Dataset demonstrate that the improved ConvGRU-CBAM achieves an F1 score of 92.86% and an AUC of 95.34%. [ABSTRACT FROM AUTHOR]

Instance segmentation is a challenging research in the field of computer vision, which combines the prediction results of object detection and semantic segmentation to provide richer image feature information. Focusing on the instance segmentation in the street scene, the real-time instance segmentation method based on SOLOv2 is proposed in this paper. First, a cross-stage fusion backbone network based on position attention is designed to increase the model accuracy and reduce the computational effort. Then, the loss of shallow location information is decreased by integrating two-way feature pyramid networks. Meanwhile, cross-stage mask feature fusion is designed to resolve the small objects missed segmentation. Finally, the adaptive minimum loss matching method is proposed to decrease the loss of segmentation accuracy due to object occlusion in the image. Compared with other mainstream methods, our method meets the real-time segmentation requirements and achieves competitive performance in segmentation accuracy. [ABSTRACT FROM AUTHOR]

Copyright of Forestry Chronicle is the property of Canadian Institute of Forestry 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.)

Oil palm trees (Elaeis guineensis) is an important species for bio-energy agribusiness. Despite the rapid growth of oil palm tree plantations mostly in tropical countries to support global demand for biofuels, problems such as diseases can reduce the productivity and survival rate of palm trees which have adverse impact to the business. Therefore. palm plantation needs regular tree counting for inventory and health monitoring. Thanks to the rapid development of remote sensing technology deep learning-based computer vision. these two intertwined technologies help to automate tree counting. Continuous improvement in this domain is expected to improve classification. In this study, YOLOv5 model was implemented for tree counting using the palm aerial imagery dataset from Papua, Indonesia. UAV images were used to classification of trees into five distinct classes, namely healthy. smallish, yellowish. mismanaged. and dead palms. We achieved average F 1 -score of 0.895 for 5 classes, which outperformed Faster R-CNN (0.706) and CNN ResNet-101 (0.493). The strength of our YOLO¥5 model is high precision for all 5 classes above 0.961. In the effort to further optimise YOLOv5. further improvements can be achieved by optimising the parameters. This was achieved using Genetic Algorithm to optimise the parameters. The final average Fl-score of this model on the five palm classes achieves 0.915. This application provides fast, robust, and accurate oil palm tree counting that can be applied elsewhere in the world. [ABSTRACT FROM AUTHOR]

Pedestrian re-identification technology enables accurate identification of individuals and is widely used in modern intelligent video surveillance systems to aid law enforcement, including criminal apprehension and locating missing persons. However, variations in lighting, background, resolution, and other imaging conditions captured by different cameras create significant challenges in pedestrian feature extraction, often leading to poor recognition accuracy. To overcome these challenges, this paper presents a Multi-resolution Feature Fusion (MRFF) method for pedestrian re-identification, based on the Pedestrian Re-identification Relational Network (RNFPR). This approach incorporates the Coordinate Attention (CA) module into the DenseNet161 network to enhance feature extraction capabilities. Improving the discriminative and recognition accuracy of features requires learning and fusing pedestrian features from multiple low-resolution images. This process enhances the expressive power of feature maps, ultimately improving pedestrian recognition performance. Additionally, this method introduces a multi-resolution feature fusion module that segments and integrates multi-resolution features from image data. This enables the model to effectively combine feature information from various resolution levels, resulting in a more comprehensive feature representation. Experimental results show that the MRFF method achieves a 1.3% increase in mean Average Precision (mAP) and a 1.0% improvement in Rank-1 accuracy on the Market1501 dataset. For the DukeMTMC-reID dataset, it provides a 0.2% increase in mAP and a 0.7% enhancement in Rank-1 accuracy. Consequently, the MRFF approach results in an overall mAP increase of 0.7% on the DukeMTMC-ReID dataset, significantly improving pedestrian gender re-identification accuracy. [ABSTRACT FROM AUTHOR]

In recent years, remote sensing object detection has become a research hotspot in computer vision tasks. However, previous approaches for remote sensing object detection often overlook the rich contextual information in images, which is crucial for accurately detecting occluded or interconnected objects using convolutional neural networks. To capture this contextual information, we propose a method called the Large Kernel Disassembling (LKD) Attention Mechanism. LKD breaks down large convolutional kernels to provide a larger receptive field to the convolutional neural networks, enabling them to capture rich contextual information in remote sensing images and enhance their performance. We employ an adaptive channel submodule and a deep convolutional spatial submodule. The adaptive channel submodule helps the network learn relationships between different channels, while the deep convolutional spatial submodule aids in extracting rich spatial features. We evaluate the proposed attention mechanism on the DIOR dataset and compare it with several recent attention mechanisms on the SSDD dataset. Experimental results demonstrate the superiority of LKD in terms of performance over other methods, validating the effectiveness of the Large Kernel Disassembling attention mechanism in remote sensing object detection tasks. [ABSTRACT FROM AUTHOR]

Steel surface defect detection poses a significant challenge in the steel industry, aiming to enhance product quality and production efficiency. Traditional mechanical and optical detection methods exhibit relatively low efficiency and poor real-time performance in detecting acceptable defects on the surface of steel strips. This paper proposes a new model named Hyper-YOLO for steel surface defect detection in steel strips. Firstly, the CSP module in the conventional YOLO backbone network is replaced with the Ghost module. The Ghost module, a lightweight convolutional module, enhances model efficiency by reducing parameter count and computational load while maintaining satisfactory performance. Secondly, researchers replace the PAFPN module in YOLO V5 in the bottleneck section with the Hyper FPN module. Hyper FPN, an improved feature pyramid network module, leverages features at different scales for multi-level feature fusion, enhancing the model's capability to detect targets at various scales. Lastly, improvements are made in the loss functions for both training and prediction stages. The α-CIoU loss function is introduced during training to substitute the original CIoU loss function, and the α-DIoU loss function is utilized during prediction instead of the original DIoU loss function. These enhanced loss functions effectively measure the accuracy and position precision of target boxes, thereby improving the detection performance of the model. Through these enhancements, the Hyper YOLO model achieves an overall performance improvement of 4.58% over the baseline model. This indicates that Hyper YOLO performs outstanding surface defect detection in steel strips, providing innovative insights for the YOLO V5 model. These improvements not only elevate the accuracy and efficiency of the model but also hold significant guidance for similar research and applications. [ABSTRACT FROM AUTHOR]

SCoralNet (based on Faster R-CNN) is a new underwater coral detection framework that has been proposed to automatically localize and identify distinct coral species in images, allowing for fast and detailed annotation. Monitoring the coverage and abundance of underwater corals typically involves the annotation and processing of large amounts of underwater coral images. However, manually annotating a large number of images is time-consuming and labor-intensive, and CNN classifiers only provide simple classification annotations without capturing the images' finer details. SCoralNet's detection performance is improved by incorporating dilated convolutions into the backbone network. To successfully capture multi-scale and multi-level information from coral targets, a neck network called NASFPN is placed between the backbone and the detecting head. Seesaw Loss is used to reduce the impact of the dataset's long-tailed distribution on SCoralNet's classifier accuracy. CIoU loss is used to optimize the bounding box regression method. During inference, Soft-NMS is applied to suppress redundant coral detection boxes. To assess SCoralNet's effectiveness, a dataset called Coral-soft was developed using real-world photos of common soft coral species from the Sanya region of China. SCoralNet outperformed the original Faster R-CNN model on the Coral-soft dataset, with a 45.68% gain in mean average precision (mAP) and a 59.2% increase in mAP75. Furthermore, SCoralNet outperformed some advanced models in terms of overall performance. [ABSTRACT FROM AUTHOR]

Objective: To examine the use of surgical intelligence for automatically monitoring critical view of safety (CVS) in laparoscopic cholecystectomy (LC) in a real-world quality initiative. Background: Surgical intelligence encompasses routine, artificial intelligence-based capture and analysis of surgical video, and connection of derived data with patient and outcomes data. These capabilities are applied to continuously assess and improve surgical quality and efficiency in real-world settings. Methods: Laparoscopic cholecystectomies conducted at 2 general surgery departments between December 2022 and August 2023 were routinely captured by a surgical intelligence platform, which identified and continuously presented CVS adoption, surgery duration, complexity, and negative events. In March 2023, the departments launched a quality initiative aiming for 75% CVS adoption. Results: Two hundred seventy-nine procedures were performed during the study. Adoption increased from39.2%in the 3 preintervention months to 69.2% in the final 3 months (P < 0.001). Monthly adoption rose from 33.3% to 75.7%. Visualization of the cystic duct and artery accounted for most of the improvement; the other 2 components had high adoption throughout. Procedures with full CVS were shorter (P = 0.007) and had fewer events (P = 0.011) than those without. OR time decreased following intervention (P = 0.033). Conclusions: Surgical intelligence facilitated a steady increase in CVS adoption, reaching the goal within 6 months. Low initial adoption stemmed from a single CVS component, and increased adoption was associated with improved OR efficiency. Real-world use of surgical intelligence can uncover new insights, modify surgeon behavior, and support best practices to improve surgical quality and efficiency. [ABSTRACT FROM AUTHOR]

The measurement of destination image from visual media such as online photography is of growing significance to destination managers and marketers who want to make better decisions and attract more visitors to their destination. However, there is no single approach with proven accuracy for doing this. We present a new approach where we fine-tune a deep learning model for a predetermined set of cognitive attributes of destination image. We then train state of the art neural networks using labelled tourist photography and test accuracy by comparing results with a ground truth dataset built for the same set of visual classes. Comparing our fine-tuned model against results which follow past approaches, we demonstrate that the pre-trained models without fine-tuning are not as accurate in capturing all of the destination image's cognitive attributes. This is, to the best of our knowledge, the first deep learning computer vision model trained specifically to measure the cognitive component of destination image from photography and can act as a benchmark for future systems. [ABSTRACT FROM AUTHOR]

This research presents a novel approach to robotic grasping by integrating an attention mechanism and advanced U-Net architectures, specifically UNet and UNet++, into the Generative Grasping Convolutional Neural Network (GG-CNN). The proposed method, Attention UNet++ GG-CNN, aims to improve the performance and accuracy of robotic grasping predictions significantly. The attention mechanism allows the model to focus on the most relevant features in the depth image, enhancing the quality of grasp predictions. The UNet and UNet++ structures, renowned for their efficiency in semantic segmentation tasks, are adapted to generate a pixel-wise grasp quality map, providing a one-to-one mapping from depth images. The UNet++ structure, with its nested and dense skip pathways, further improves the model's ability to capture and propagate low-level details to higher-level features. This approach overcomes the limitations of traditional deep learning grasping techniques by avoiding discrete sampling of grasp candidates and reducing computation times. Preliminary results indicate that the Attention UNet++ GG-CNN significantly improves the accuracy and performance of robotic grasping,with the Attention UNet++ GG-CNN achieving Intersection over Union (IoU) scores up to 95.86%, and average performance metrics across 25%, 50%, and 75% thresholds reaching as high as 88.29%, confirming its potential for more effective and reliable robotic manipulation in various settings, paving the way for more efficient and reliable robotic manipulation in dynamic environments. [ABSTRACT FROM AUTHOR]

Increasing deformity of the lower extremities, as measured by the hip-knee-ankle angle (HKAA), is associated with poor patient outcomes after total hip and knee arthroplasty (THA, TKA). Automated calculation of HKAA is imperative to reduce the burden on orthopaedic surgeons. We proposed a detection-based deep learning (DL) model to calculate HKAA in THA and TKA patients and assessed the agreement between DL-derived HKAAs and manual measurement. We retrospectively identified 1,379 long-leg radiographs (LLRs) from patients scheduled for THA or TKA within an academic medical center. There were 1,221 LLRs used to develop the model (randomly split into 70% training, 20% validation, and 10% held-out test sets); 158 LLRs were considered "difficult," as the femoral head was difficult to distinguish from surrounding tissue. There were 2 raters who annotated the HKAA of both lower extremities, and inter-rater reliability was calculated to compare the DL-derived HKAAs with manual measurement within the test set. The DL model achieved a mean average precision of 0.985 on the test set. The average HKAA of the operative leg was 173.05 ± 4.54°; the nonoperative leg was 175.55 ± 3.56°. The inter-rater reliability between manual and DL-derived HKAA measurements on the operative leg and nonoperative leg indicated excellent reliability (intraclass correlation (2,k) = 0.987 [0.96, 0.99], intraclass correlation (2, k) = 0.987 [0.98, 0.99, respectively]). The standard error of measurement for the DL-derived HKAA for the operative and nonoperative legs was 0.515° and 0.403°, respectively. A detection-based DL algorithm can calculate the HKAA in LLRs and is comparable to that calculated by manual measurement. The algorithm can detect the bilateral femoral head, knee, and ankle joints with high precision, even in patients where the femoral head is difficult to visualize. [ABSTRACT FROM AUTHOR]

Camber in hot-rolled plates significantly impacts product quality and rolling process stability, making accurate camber prediction crucial. However, it is challenging to measure asymmetric factors impacting camber in real production, hindering the ability of current models to predict and analyze the overall camber of rolled plates. This study proposes a method that combines asymmetric fluctuations of measurable variables with deep learning to predict camber. We developed a data analysis platform for plate processing and constructed a camber dataset using machine vision technology. During the modeling phase, a hot-rolled plate camber prediction model, BiLSTM-AM-Seq2Seq, was proposed, integrating a bidirectional long short-term memory network, an attention mechanism, and a sequence-to-sequence model. Additionally, an improved scheduled sampling method was also introduced to steer model training. Model performance was evaluated with real-world production data, demonstrating superior accuracy and stability. Specifically, the model achieved a mean absolute error of 12.29 mm and a root mean square error of 17.32 mm. Consequently, this model meets the demands of practical production and addresses the need for overall camber prediction of hot rolled plates. • Data-driven prediction model for overall camber shape of rolled plates is proposed. • Modeling using fluctuations of observable variables to address the absence of key asymmetric variables in real production. • Data analysis platform was developed and real data were used to demonstrate the superiority of the proposed model. [ABSTRACT FROM AUTHOR]

Video segmentation and object tracking are critical tasks in computer vision with applications spanning surveillance, autonomous driving, and interactive media. Traditional methods often struggle with the dynamic nature of video data, where object occlusions, variations in illumination, and complex motion patterns present significant challenges. Existing segmentation and tracking systems frequently suffer from inaccuracies in handling real-time video sequences, particularly in distinguishing and tracking multiple overlapping objects. The limitations of current models in addressing these issues necessitate the development of more advanced techniques that can effectively manage dynamic scenes and improve tracking accuracy. To address these challenges, we propose an advanced machine learning technique, AI-Enhanced TrackSegNet, which integrates deep learning with novel attention mechanisms for improved video segmentation and object tracking. Our method utilizes a combination of Convolutional Neural Networks (CNNs) for feature extraction and Long Short-Term Memory (LSTM) networks for temporal sequence modeling. We introduce an attention-based mechanism to dynamically focus on relevant features, enhancing the model's ability to handle occlusions and varying object appearances. The model was trained on a diverse dataset of video sequences, incorporating both synthetic and real-world footage. The AI-Enhanced TrackSegNet demonstrated significant improvements in performance compared to existing techniques. Our method achieved an average Intersection over Union (IoU) score of 86.7% for segmentation and a tracking precision rate of 91.3% on the MOT17 benchmark dataset. These results represent a 10.2% improvement in IoU and a 7.5% increase in tracking precision compared to state-of-the-art methods. The model also exhibited enhanced robustness in complex scenes, handling occlusions and motion variations with greater accuracy. [ABSTRACT FROM AUTHOR]

Defect detection is vital in the nonwoven material industry, ensuring surface quality before producing finished products. Recently, deep learning and computer vision advancements have revolutionized defect detection, making it a widely adopted approach in various industrial fields. This paper mainly studied the defect detection method for nonwoven materials based on the improved NanoDet-Plus model. Using the constructed samples of defects in nonwoven materials as the research objects, transfer learning experiments were conducted based on the NanoDet-Plus object detection framework. Within this framework, the Backbone, path aggregation feature pyramid network (PAFPN) and Head network models were compared and trained through a process of freezing, with the ultimate aim of bolstering the model's feature extraction abilities and elevating detection accuracy. The half-precision quantization method was used to optimize the model after transfer learning experiments, reducing model weights and computational complexity to improve the detection speed. Performance comparisons were conducted between the improved model and the original NanoDet-Plus model, YOLO, SSD and other common industrial defect detection algorithms, validating that the improved methods based on transfer learning and semi-precision quantization enabled the model to meet the practical requirements of industrial production. [ABSTRACT FROM AUTHOR]

This survey article provides an expose of the relative literature with a specific focus in conveyor belt systems incorporation of artificial intelligence (AI). This survey article describes the belt condition and its prognostics based on IoT, performance analysis, visualization, and force mailing. The review is based on over 79 articles of peer-reviewed journals published in the last five years of the study and focused on the enhancement of performance and safety of conveyor belt systems for manufacturing, mining, and logistics industries applying advanced AI techniques using DL models. The AI technologies to be investigated are majors of the ML algorithms to be used in the detection of faults and prediction of failures, CV systems to be used in real-time identification of defects on the assets and IoT systems to be used in the collection and processing of data. From the survey, it is seen that the integration of these set of possibilities of AI enhances the competency in the areas of accurate fault detection; superior control and computer based intelligent operation of the material handling than the aspect of monitoring the fan conveyor. Innovation involves some concepts that include the following; belt tear prediction models using neural networks with more than 95 percent certainty for the real time prediction of belt tears, computer vision, for the real time identification of surface issues, IoT that can reduce the system's unplanned time by at least 30 percent. It also describes the current state of affairs when it comes to data quality problem, explanation of the algorithms used and the procedure of scaling up the already existing systems. Last but not least, it offers key and precise recommendations for the further research on the multiple levels of intelligence in AI systems as well as the Edge AI intelligent decisions, the Reinforcement Learning intelligent control, and AI with other emerging technologies; Digital Twin. Finally, it might be mentioned that, concerning the survey made, it is possible to state how the conveyer belt system may be altered with the competent usage of the AI in various fields for making performance, reliability, as well as security improvements. [ABSTRACT FROM AUTHOR]

This article elucidates the concept of large model technology, summarizes the research status of large model technology both domestically and internationally, provides an overview of the application status of large models in vertical industries, outlines the challenges and issues confronted in applying large models in the oil and gas sector, and offers prospects for the application of large models in the oil and gas industry. The existing large models can be briefly divided into three categories: large language models, visual large models, and multimodal large models. The application of large models in the oil and gas industry is still in its infancy. Based on open-source large language models, some oil and gas enterprises have released large language model products using methods like fine-tuning and retrieval augmented generation. Scholars have attempted to develop scenario-specific models for oil and gas operations by using visual/multimodal foundation models. A few researchers have constructed pre-trained foundation models for seismic data processing and interpretation, as well as core analysis. The application of large models in the oil and gas industry faces challenges such as current data quantity and quality being difficult to support the training of large models, high research and development costs, and poor algorithm autonomy and control. The application of large models should be guided by the needs of oil and gas business, taking the application of large models as an opportunity to improve data lifecycle management, enhance data governance capabilities, promote the construction of computing power, strengthen the construction of "artificial intelligence + energy" composite teams, and boost the autonomy and control of large model technology. [ABSTRACT FROM AUTHOR]

During the packaging process, it is essential to detect the steel stamping characters inside the box to identify any missing or repeated characters. Currently, manual detection suffers from low efficiency and a high false detection rate. To address these challenges, a steel stamping character recognition and analysis system based on machine vision has been developed. The enhanced YOLOv7 detection method was employed for character identification, complemented by a statistical analysis approach to achieve automated judgment and detection. To address the issue of size disparity between large and small characters, a small size anchor box and a larger detection head were integrated. Furthermore, modifications were made to the output structure of the YOLOv7 prediction network to enhance multi-scale detection capabilities. The inclusion of the location attention convolution module bolstered global feature extraction, thereby enhancing the detection accuracy of similar characters. Moreover, the utilization of a hash table was used to improve the efficiency of mapping steel stamping character recognition sequences. The experimental results demonstrate that the enhanced model achieves an accuracy of 99.83%, with a processing efficiency of 10.5 ms per single frame. These findings align with the performance criteria for automatic recognition and analysis of steel stamping characters. [ABSTRACT FROM AUTHOR]

For a variety of production and development processes, there is a need for precise and reliable measurements of rotating machinery or objects. In this article, an alternative approach is introduced to assess both rotational and translational movements in a rotor by employing the technique known as "visual encoder." This methodology combines principles of computer vision with visual components, operating analogously to conventional optical encoders. Non-contact measurements provided by visual encoders offer several advantages, including the ability to withstand free movements and oscillations along the rotation axis, as well as application in hostile environments, such as high-temperature conditions that might pose challenges for conventional measurement methods. The proposed method incorporates translational motion tracking into rotational velocity measurement using cost-effective conventional equipment. Experimental tests demonstrate that the developed system exhibited high precision and robustness under various operating conditions, successfully operating even at rotational frequencies close to the sampling rate. The results validate the developed technique as a viable alternative for measuring rotation and translational movement in rotor applications. [ABSTRACT FROM AUTHOR]

Stress is a state of mental tension, which helps us to cope with challenges in our life. It makes us progressive when it is positive, but excessive negative stress that perseveres for a long time leads to a state of depressiveness. Longer stressed stage of a human being changes the size, functionality and frequency of response of many internal and external body parameters. By applying computer vision techniques, these changes of body parameters can be tracked to get useful information about the mental stress for a stress affected person. Many studies show the pupil diameter varies significantly with the effect of stress. Our work is based on the study of variation of pupil diameters of stress affected and not affected university students. With the application of different supervised machine learning algorithms, we have observed that the pupil dilates more in case of stress affected students than non-stressed students. We have also found that the pupils of the students dilates more when they were in positive emotional states than their negative emotional states. This work will be helpful for researchers who are working in the field of emotion detection and recognition and affective disorder analysis. [ABSTRACT FROM AUTHOR]

While new buildings continue to emerge in the process of urbanization, historical buildings, as valuable legacies carrying national historical memory, play an important role in the urban landscape. Previous studies have shown that color harmony is a crucial factor in coordinating urban landscapes. However, the evaluation of color harmony in historic areas and buildings lacks effective quantitative standards, often overlooking factors such as complementary color harmony and the compatibility of analogous colors. This study aims to build a new method to evaluate the color harmony of historical buildings through street view technology, semantic segmentation algorithms, quantification of color harmony methods based on image property detection and classification, questionnaire verification, and takes Shanghai's historical buildings as an example to explore. Our study categorizes six types of color harmony indexes for Shanghai street-facing historic buildings into three levels, with the top tier serving as a benchmark for excellence and the lowest tier highlighting areas in need of urban environmental improvement. This study uniquely considers color compatibility within hue ranges and expanded relationship types like complementary harmony. This approach, applicable to cities globally, offers practical tools for urban planners and conservators in managing and preserving historic areas and buildings. [ABSTRACT FROM AUTHOR]

The palm oil industry has always been a pivotal sector on the world's vegetable crop industry, especially for southeast Asian countries such as Indonesia and Malaysia. These fast-growing sectors has led to challenges for assessing oil palm fruit maturity. In order to overcome these obstacles, this research suggests an automating the detection of oil palm fruit maturity through the use of computer vision. The six maturity clusters for oil fruit images are Rotten, Underripe, Unripe, Overripe, and Empty Bunch. Several different types of YOLOv8 models, including YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x, are evaluated in this study. The YOLOv8m emerges as the best option, which shows notable gains in accuracy metrics at the same time as processing speeds. For example, YOLOv8m processes at an average speed of about 14.68 iterations per second and achieves top-1 and top-5 accuracy of 0.9885 and 0.998, respectively. [ABSTRACT FROM AUTHOR]

The escalating prevalence of traffic congestion in Malang, which has worsened over time, has resulted in increased inconvenience and traffic-related issues. As a solution, we propose the development of an intelligent transportation system based on traffic flow estimation. While this endeavor presents challenges, advancements in traffic data accessibility and deep learning algorithms, including You Only Look Once (YOLO), R-CNN based ResNet101, Inception V2, and others, with YOLOv7 demonstrating superior performance, continue to enhance traffic flow estimation. Consequently, this study aims to describe and assess the YOLOv7 model's suitability for estimating traffic flow. To achieve this, we propose a traffic monitoring system utilizing the YOLOv7 model in conjunction with a SORT filter, capable of tracking vehicle count and average speed using a custom dataset comprising car images and CCTV footage of Malang's traffic. Subsequently, the number and average speed of cars passing on the road were computed by delineating two virtual boxes. According to the findings, our YOLOv7 system exhibited higher average precision, achieving 61.3%, 78.6%, and 62.1% for each test, and boasted a faster average inference time of 90.67 seconds, meeting the requirements of an intelligent transportation system. Nevertheless, with an average percent deviation in vehicle count exceeding 30% compared to the pre-trained model, vehicle detection is suboptimal, consequently impacting average vehicle speed detection adversely. Hence, our custom model necessitates further optimization and training before practical implementation in real-world traffic estimation. [ABSTRACT FROM AUTHOR]

The widespread implementation of surveillance systems on construction sites has led to the accumulation of vast amounts of visual data, highlighting the need for an effective semantic analysis methodology. Natural language, as the most intuitive mode of expression, can significantly enhance the interpretability of such data. The adoption of multi-modality models promotes the interaction between surveillance video and textual data, thereby enabling managers to swiftly comprehend on-site dynamics. This study introduces a Visual Question Answering (VQA) approach for the construction industry and presents a specialized dataset to address the unique requirements of on-site management. Utilizing a Vision Transformer (ViT) architecture, the proposed model conducts feature extraction, fusion and interaction between visual and textual features. An additional projection layer is added to establish a transfer learning strategy that is optimized for construction site data. This novel approach facilitates rapid alignment of visual and language features in the model and is validated through ablation studies. The proposed approach achieves a testing accuracy of 83.8%, effectively converting image data from construction sites into natural language descriptions that enhance the analysis of construction processes. Compared to existing methods, this approach does not rely on object detection and allows for the direct extraction of deep-level semantic information from the on-site images. This study further discusses the feasibility of applying VQA within the architecture, engineering and construction (AEC) industry, examines its limitations, and offers suggestions for viable future directions of development. [ABSTRACT FROM AUTHOR]

This research study offers a novel method that uses deep learning to improve the processing of scrap materials segregation and management. There is no need to stress the importance of effective scrap management. The data was obtained from Kaggle and reorganized into forty-three classes. Classes with low data were improved by downloading additional images from Google search and taking photo shots. This model aims to identify and classify scrap materials (plastic, glass, etc.) composed of 43 types of scrap items. Convolutional Neural Networks (CNNs) with computer vision, a cutting-edge technology, are used with particular emphasis on methods investigated for comparison, including Logistic Regression, AdaBoost, and Random Forest for classifying material. The model's training phase shows an astounding 95% accuracy in classifying waste items. At the same time, the results of testing data sets for different objects show different accuracies ranging from 80 - 90%. Ultimately, CNN and logistic regression were found to be the promising option because of their sophisticated object-detecting skills, even though Logistic Regression gave good competition because of few evaluation parameters. Additionally, it goes a step further by recommending appropriate recycling routes for the scrap that can be recycled. The concept speeds up the collection and transportation of recyclable waste by automating segregation. The operational difficulties that companies involved in managing discarded materials must deal with are directly addressed by this paper. Importantly, this practical strategy may provide a solid response to India's urgent scrap management challenges. The model is in line with circular economy principles and sustainable business practices. This research extends its advantages to supply chain management and enterprises by fostering sustainability. Considering the global trend toward sustainability, the appeal of implementing such ecologically friendly methods not only addresses the current but also plots a road for future prosperity. [ABSTRACT FROM AUTHOR]

With the rapid development of Internet of Things technology, vehicle-based entertainment systems are developing towards a more natural, faster, and automated direction. To achieve automatic control of vehicle-based music, based on machine vision, the music automation control method is designed by recognizing human gestures. The improved gesture motion trajectory point localization method, and enhanced recognition method based on the HMM are proposed to control vehicle-based music. The research results indicated that the method proposed in this research had a recognition rate of 95.3% for non-similar gestures. The recognition rate for similar gestures reached 94.56%. This indicates that the designed method can accurately and automatically control in car music equipment It is expected that this research can provide some reference value for the current research on vehicle-based music automated control. [ABSTRACT FROM AUTHOR]

The advancements in drone technologies, digital imaging, computer vision techniques, and the liberalized laws related to drone flying have opened up drone-based applications such as the delivery of supplies, search and rescue, aerial surveillance, and so on. The drones, especially the nano/micro/small drones, may be mounted with only low-resolution camera(s) due to their maximum takeoff weight limitations. The low-resolution images generated by the cameras, if used for landing, can result in faulty detection unless the photos are taken from a very close distance to the point of interest. Detection and recognition of the point(s) of interest as early as possible is required to ensure sufficient response time for safe maneuvering. Hence, the images are to be captured at greater heights or distances from the point(s) of interest, and obtaining the high-resolution images from the captured low-resolution images is crucial. The High Resolution (HR) and the Low Resolution (LR) image pairs for training super-resolution models in the works presented in literature are generated using two different cameras or the HR images are captured by the camera and LR images are generated by degrading the HR images. As both methods are not appropriate for small/ micro/nano category drones, we propose a novel method based on Ground Sampling Distance (GSD) to capture the LR and HR images. In this paper, we have presented the designed methodology for the creation of a dataset using drone-mounted cameras covering a broad spectrum of views of the target(s) suitable for training and testing of the Single Image Super-Resolution (SISR) models. We also present a methodology for selecting an appropriate target for imaging that enables the visual quality assessment of the developed super-resolution model. [ABSTRACT FROM AUTHOR]

The applications of digital images is widespread in various fields such as medical diagnosis, film production, intelligent transportation, security monitoring, virtual reality and computer vision. The security of image transmission is a widely discussed topic, and ensuring the protection of digital images has become a focal point in Internet security communication. The selection of an algorithm with strong encryption performance holds significant theoretical and practical importance for image transmission. This paper takes the pseudo-random sequence as a starting point and primarily utilizes scrambling operations, the Mersenne rotation algorithm, and operations based on chaotic sequences to explore image encryption algorithms. Firstly, it introduces the encryption principles of these three methods, followed by respective application cases and implementation using MATLAB software. Pseudo-random sequences are employed as tools in all three methods, reo rdering pixel positions according to their order in the pseudo-random sequence during scrambling operation. In the Mersenne rotation algorithm, pseudo-random sequences are used to generate random number seeds. Chaotic sequences are utilized in combination with scrambling operations to achieve image encryption through generating pseudo-random sequences. Finally, after analyzing information entropy, histograms, encryption/decryption times, and adjacent pixel correlations, we select the optimal image encryption algorithm that provides better digital image encryption effects and higher security levels for each application case along with code implementations. [ABSTRACT FROM AUTHOR]

Technologies such as big data and computer vision are increasingly being applied in trade logistics, and their integration provides strong support for achieving modernization and intelligence in the logistics industry. This article utilizes various sensors, global positioning systems, radio frequency identification and other technologies to obtain real-time information on the location, temperature, and transportation status of goods. Video surveillance and image recognition methods can be used, combined with computer vision technology, to monitor the loading and unloading process of goods and the condition of transportation vehicles in real time. This article studies an intelligent monitoring algorithm that achieves anomaly detection, warning prompts, and ensures the safety and smoothness of transportation. Exploration of transportation path optimization in batch 1 experimental group (big data analysis) saved 3 hours of time and cost by 500 yuan. This article explores the intelligent monitoring and management of trade logistics based on big data and computer vision technology, which can help improve transportation efficiency and reduce costs. [ABSTRACT FROM AUTHOR]

Precise positioning of solder joints in body-in-white is an important step in realizing intelligent quality inspection of solder joints. In conventional factory environments, affected by factors such as light, surface oil, and electrode cap wear, the burnt black outline layer at the solder joints is discontinuous. This makes the gray scale distribution of the internal area of the solder joint more uncertain, and the accuracy of conventional visual circular positioning detection methods is not high. This paper uses a combination of texture feature analysis and regional statistics to locate the solder joints of the body-in-white. Experiments have proved that this method This method can show high resistance to light interference and achieve precise positioning of solder joints in complex factory environments. [ABSTRACT FROM AUTHOR]

Machine vision is based on computer, through the comprehensive utilization of a series of advanced theoretical knowledge such as image acquisition equipment, digital processing technology and related algorithms. It can quickly and accurately extract the required features from the input data and describe them, which plays an important role in modern industrial production. Machine vision can intuitively express the image information that people are exposed to in a certain way, and process and process it accurately and quickly. This paper aims to use machine vision to quickly obtain spatial position, size and other data, combined with artificial intelligence genetic algorithm, particle swarm algorithm, intelligent plane optimization. This paper mainly uses the experimental simulation method and the comparison method to study the optimization method of the system layout, puts forward two improvement schemes of the storage system layout, and makes a comparative analysis of their space loss parameters. The experimental results show that in terms of total space loss, the value of improved scheme 1 is 0.741, which is higher than that of improved scheme 2. [ABSTRACT FROM AUTHOR]

Currently, a growing body of research focuses on addressing problems using computer vision libraries and artificial intelligence tools. The predominant approaches involve employing machine and deep learning models of artificial neural networks to recognize gestures in the Kazakh Sign Alphabet (KSA) via supervised and deep learning techniques for sequential data processing. Pattern recognition in this context involves identifying an object within an image, where the object can be abstract and vary in shape. We have chosen to investigate the field of gesture recognition, specifically. For recognizing Kazakh Sign Language (KSL), the initial step involves mastering the KSA. Training a neural network to recognize KSL necessitates the collection of datasets in the form of images depicting hand gestures. In this research, prominent hand gesture recognition models such as the Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Support Vector Machine (SVM) were analyzed. These models differ in their methodologies, processing times, and training data requirements. A significant aspect of this study is the application of unsupervised and supervised learning techniques including CNN, LSTM, and SVM. The experiments yielded diverse results when training neural networks for recognizing gestures in Kazakh sign language based on the dactyl alphabet. This article provides a comprehensive overview of each method, their specific purposes, and their effectiveness in terms of performance and training. Numerous experimental outcomes were documented in a table, showcasing the accuracy of recognizing each gesture. Additionally, specific hand gestures were tested in front of a camera to identify the gesture and display the result on the screen. A notable feature was the use of mathematical formulas and functions to elucidate the operating principles of the machine learning methods, as well as the logical structure and design of the LSTM model. [ABSTRACT FROM AUTHOR]

This paper introduces research aimed at developing and implementing an intelligent passenger counting system in public transportation. The goal is to enhance fare evasion control and reduce financial losses for transit operators by automatically monitoring the entering and exiting of passengers. Utilizing the advanced capabilities of YOLO and DeepSORT algorithms, the system demonstrates high accuracy in detecting and tracking passengers, even in congested scenarios. Practical experiments have highlighted several challenges, including real-time frame processing rates, single-board computer performance, and the accuracy of passenger counts. It was discovered that the selection and positioning of USB cameras significantly influence system efficiency. Notably, changing the IP cameras to USB cameras significantly improves the processing speeds. A dataset containing images was created to improve the YOLO model's accuracy in detecting passengers. This dataset was particularly relevant considering the winter season, where identifying passengers could be challenging due to their bulky outerwear. To upgrade the performance of the DeepSORT system, the YOLO model was also improved with the pre-designed dataset to focus solely on the heads. This trained model achieved a recognition accuracy of 91% during peak hours. During periods of lighter traffic, when there were fewer passengers, the accuracy increased further due to reduced interference and higher precision. To enhance real-time processing efficiency, this research incorporates the status of public transport doors as a key parameter. By optimizing the number of frames forwarded to the YOLO algorithm for processing, we ensure that the system maintains accuracy without compromising on speed. Future development will focus on incorporating depth-sensing cameras and developing advanced data analysis algorithms for real-time processing on single-board computers. This will further enhance the precision of passenger counts during high-traffic periods. Expected benefits of this project include more efficient transport route planning, better management of passenger flows, improved rider satisfaction, and more effective measures against fare evasion. [ABSTRACT FROM AUTHOR]

Falling is a significant threat for old people and disabled persons. Therefore, it is necessary to develop an automated fall detection system to make them more independent and to mitigate the physical and psychological consequences of falls. Such a system can lower the death rate among dependent people who live alone. This paper presents a review of vision-based techniques, algorithms and databases used in detecting falls. Recently proposed systems are analyzed and compared. The literature review underscores the promising potential of camera-based fall detection systems and highlights areas for further research and development. [ABSTRACT FROM AUTHOR]

The use of visual rationality in planning and designing indoor 3D spaces can effectively enhance the quality of environmental art design. For the rational planning and design of 3D spaces, it is essential to assign attribute values to the rational evaluation index of 3D space planning, determine the weight vector of rational evaluation factors for indoor 3D space planning, and accomplish the visual rationality planning and design of indoor 3D spaces. 3D reconstruction is one of the primary objectives and focal points of machine vision research. The principle is to extract 3D information from the 2D image captured by the machine’s sensors in order to reconstruct a 3D scene model. The traditional method constructs the evaluation index system for indoor 3D space planning to determine the weight of the evaluation index for alternative points. However, it overlooks the weight vector of the evaluation factors, leading to low accuracy in rational planning. This paper optimizes the effect of structured light reconstruction using photometric stereo vision. Firstly, the text delves into the impact of light sources on the 3D reconstruction effect. Subsequently, a fusion algorithm combining 3D and structured light reconstruction methods is introduced, leveraging artificial intelligence (AI) technology. Experimental results show that this algorithm has a superior 3D reconstruction effect. [ABSTRACT FROM AUTHOR]

A conventional codec aims to increase the compression efficiency for transmission and storage while maintaining video quality. However, as the number of platforms using machine vision rapidly increases, a codec that increases the compression efficiency and maintains the accuracy of machine vision tasks must be devised. Hence, the Moving Picture Experts Group created a standardization process for video coding for machines (VCM) to reduce bitrates while maintaining the accuracy of machine vision tasks. In particular, in‐loop filters have been developed for improving the subjective quality and machine vision task accuracy. However, the high computational complexity of in‐loop filters limits the development of a high‐performance VCM architecture. We analyze the effect of an in‐loop filter on the VCM performance and propose a suboptimal VCM method based on the selective activation of in‐loop filters. The proposed method reduces the computation time for video coding by approximately 5% when using the enhanced compression model and 2% when employing a Versatile Video Coding test model while maintaining the machine vision accuracy and compression efficiency of the VCM architecture. [ABSTRACT FROM AUTHOR]

Background: Early-life pain is associated with adverse neurodevelopmental consequences; and current pain assessment practices are discontinuous, inconsistent, and highly dependent on nurses' availability. Furthermore, facial expressions in commonly used pain assessment tools are not associated with brain-based evidence of pain. Purpose: To develop and validate a machine learning (ML) model to classify pain. Methods: In this retrospective validation study, using a human-centered design for Embedded Machine Learning Solutions approach and the Neonatal Facial Coding System (NFCS), 6 experienced neonatal intensive care unit (NICU) nurses labeled data from randomly assigned iCOPEvid (infant Classification Of Pain Expression video) sequences of 49 neonates undergoing heel lance. NFCS is the only observational pain assessment tool associated with brain-based evidence of pain. A standard 70% training and 30% testing split of the data was used to train and test several ML models. NICU nurses' interrater reliability was evaluated, and NICU nurses' area under the receiver operating characteristic curve (AUC) was compared with the ML models' AUC. Results: Nurses weighted mean interrater reliability was 68% (63%-79%) for NFCS tasks, 77.7% (74%-83%) for pain intensity, and 48.6% (15%-59%) for frame and 78.4% (64%-100%) for video pain classification, with AUC of 0.68. The best performing ML model had 97.7% precision, 98% accuracy, 98.5% recall, and AUC of 0.98. Implications for Practice and Research: The pain classification ML model AUC far exceeded that of NICU nurses for identifying neonatal pain. These findings will inform the development of a continuous, unbiased, brain-based, nurse-in-the-loop Pain Recognition Automated Monitoring System (PRAMS) for neonates and infants. [ABSTRACT FROM AUTHOR]

In den letzten Jahren hat die digitale Revolution auch die Prävention und Rehabilitation von Sportverletzungen rasant erfasst. Für die Prävention spielt die Digitalisierung bereits durch die Integration verschiedener Technologien und Ansätze eine wichtige Rolle. Dazu gehören Wearables, Sensoren und Tracking-Plattformen, virtuelles Training und Simulation sowie Big Data und künstliche Intelligenz (KI). Die Digitalisierung ermöglicht mit diesen Methoden einen ganzheitlichen Ansatz zur Verletzungsprävention im Sport, indem Echtzeitdaten, Analysetools und innovative Technologien genutzt werden, um potenzielle Risiken zu erkennen und um umgehend präventive Maßnahmen zu ergreifen. Auch in der Rehabilitation von Sportverletzungen spielt die Digitalisierung heute eine große Rolle. Mobile Anwendungen, in die KI, Sensoren oder virtuelle Realität integriert werden können, werden in der sportmedizinischen Rehabilitation bereits klinisch eingesetzt. Diese Apps können den Rehabilitationsprozess unterstützen, indem sie Übungsprogramme vorschlagen, Feedback über Sensoren geben und den Rehabilitationsprozess mithilfe von KI an die aktuellen Fähigkeiten des Patienten anpassen. Zusammenfassend lässt sich sagen, dass die Digitalisierung zweifellos das Potenzial hat, unser bisheriges Verständnis von Prävention und Rehabilitation zu revolutionieren. Um die Risiken dieser neuen Technologien einschätzen zu können, sind jedoch weitere Forschungsarbeiten erforderlich. In recent years, the digital revolution has rapidly affected the prevention and rehabilitation of sports injuries. For prevention, digitalization plays an important role by integrating various technologies and approaches. These include wearables, sensors and tracking platforms for monitoring, motion capture for biomechanical analysis, virtual training and simulation, and big data and artificial intelligence (AI) for identifying patterns and relationships that indicate risk of injury. Digitalization therefore enables a holistic approach to injury prevention in sport by using real-time data, analysis tools and innovative technologies to identify potential risks and take appropriate measures. Digitalization also plays a major role in the rehabilitation of sports injuries today. Mobile applications in which AI, sensors or virtual reality can be integrated are already used clinically in sports medicine rehabilitation. These apps can support the rehabilitation process by suggesting exercise programs, providing feedback via sensors and using AI to adapt the rehabilitation process to the patient's current abilities. In summary, digitalization undoubtedly has the potential to revolutionize the way we have previously understood prevention and rehabilitation. However, further research is necessary to be able to assess the risks of these new technologies. [ABSTRACT FROM AUTHOR]

It is the refrigerator that is frequently used in homes. Freezing and refrigerating temperature sensors are significant temperature-sensing components in the refrigerator. However, it is simple to insert the temperature sensor wires erroneously during the manufacturing process. Currently, there is no effective method to detect the correctness of the wiring harnesses. We propose a novel detection method that combines machine vision and resistance analysis for the wiring harnesses of refrigerator temperature sensors. We utilized machine vision technology to determine the color and insertion position of each wire in the wiring harnesses. We distinguish the types of sensors by analyzing the resistance range of the temperature sensors. We combine two approaches to detect the wiring harnesses connecting the temperature sensors. The experimental results demonstrate that the proposed method can effectively detect the correctness of the wiring harnesses connecting the temperature sensors, and the accuracy reaches 99.19%. The proposed method has a good generalization in other electrical wiring harness detection problems. [ABSTRACT FROM AUTHOR]