Research on Defect Detection Method of Drainage Pipe Network Based on Deep Learning

As the core component of electronic products, the quality of PCB (print circuit board) directly affects the performance and reliability of the entire product. The traditional PCB defect detection methods mainly rely on manual visual inspection and image processing technology, which have problems such as low efficiency and easy errors and cannot meet the needs of large-scale production. The PCB defect detection method based on deep learning and object detection technology has gradually become mainstream. This article focuses on the research of PCB defect object detection method based on deep learning. Therefore, the paper conducts relevant intensive study on PCB defect detection, aiming to develop a model that is accurate, compact, and suitable for real-time detection. Specific improvements include using C2F_The ViT module replaces the original C2F, and the purpose is to facilitate the model to comprehensively consider the local details of PCB defects and the global contextual semantic information. Introduces multi-head self-attention mechanism (MHSA), adds detection heads in the p2 layer, and improves defect detection accuracy and location accuracy. Extract the target semantic information of small defects and suppress the local background information at the same time. The experimental results indicate that the detection algorithm can improve detection accuracy. Compared with the original YOLOv8, the improved model has a 4.9% and 7.9% improvement in map50 and map50-95, respectively.

Normally functioning and complete printed circuit board (PCB) can ensure the safety and reliability of electronic equipment. PCB defect detection is extremely important in the field of industrial inspection. For traditional methods of PCB inspection, such as contact detection, are likely to damage the PCB surface and have high rate of erroneous detection. In recent years, methods of detection through image processing and machine learning have gradually been put into use. However, PCB inspection is still an extremely challenging task due to the small defects and the complex background. To solve this problem, a lightweight one-stage defect detection network based on dual attention mechanism and Path Aggregation Feature Pyramid Network (PAFPN) has been proposed. At present, some methods of defect detection in industrial applications are often based on object detection algorithms in the field of deep learning. Through comparative experiments, compared with the Faster R-CNN and YOLO v3 which are usually used in the current industrial detection, the inference time of our method are reduced by 17.46 milliseconds (ms) and 4.75 ms, and the amount of model parameters is greatly reduced. It is only 4.42 M, which is more suitable for industrial fields and embedded development systems. Compared with the common one-stage object detection algorithm Fully Convolutional One-Stage Object Detection (FCOS), mean Average Precision (mAP) is increased by 9.1%, and the amount of model parameters has been reduced by 86.12%.

Defects in key components of railway vehicles caused by long-running is inevitable. Automated data mining for defect detection is significant for operation and maintenance in railway industry. A huge number of inspection images put pressure on manual analysis. In order to reduce the cost of manual inspection, this paper proposes an automated data mining method for defect detection named multi-channel detection framework (MCDDF). This method regards defect detection task as two stages pipeline, from coarse to fine process. Including one channel for locations and classifications of key components, and the other for judging defects categories of key components. Experiments on high-speed train defect dataset and some frequently used datasets, from various domains, verifies effectiveness of MCDDF over previous approaches.

Advances in deep learning and computer vision have revolutionized object detection, enabling real-time and accurate object recognition. These object detection technologies can potentially transform accessibility solutions, especially for individuals with visual impairments. This study aims to enhance accessibility and environment-effective interaction for individuals with visual disabilities by detecting and naming objects in real-world environments. This study examines and optimizes the potential of a set of developed deep learning models, including YOLOv8L, YOLO11x, and Faster region-based convolutional neural network (R-CNN) with seven backbone models for multi-class object detection to enhance object recognition and provide auditory feedback; these models aim to bridge the gap between the visually impaired and their surroundings. In addition, we attempt to propose a system that translates detections into audible descriptions, empowering individuals to navigate and interact with the world independently by integrating object detection with text-to-speech (TTS) technology. The models leverage Arabic-translated PASCAL VOC 2007 and 2012 datasets, with performance evaluated through precision, recall, and mean average precision (mAP). The results revealed that YOLO11x achieves the highest mAP of 0.86, followed by YOLOv8L with an mAP of 0.83. Faster R-CNN with EfficientNet-B3, HRNet-w32, and MobileNetV3-Large showed the highest accuracy among other backbones with 79%, 78%, and 75%, respectively. The study proves the efficacy of deep learning models in accessibility applications as assistive technologies for individuals with visual impairments and highlights opportunities for future development.

In the rapid development of deep learning, YOLO, as the first popular single-stage object detection model, has sparked an innovative storm in the computer vision community with its remarkable architecture and innovative concepts, marking a significant leap forward in object detection technology. Today, it is not only regarded as a milestone in this field but also sets an unparalleled example in the pursuit of the perfect fusion of detection speed and accuracy. YOLO has been widely applied in various fields such as agriculture, industry, pedestrian detection, and more. This research project will first introduce traditional object detection methods, then analyze object detection based on deep learning, and subsequently elaborate on the fundamental concepts of YOLO. It will systematically sort through the YOLO family and its significant improvements. Finally, based on different improvement strategies or application scenarios, the YOLO algorithm will be systematically classified and summarized.

Recently, object detection technology using deep learning has been rapidly developed. In this paper, we propose a method of generating trainset by automatically merging objects to recognize occluded objects using deep learning. Since the objects are merged automatically by using various distances and angles among objects, generated trainset can cover a lot of different cases while saving time and labor compared to manual labeling. The proposed system achieves a correct detection rate of 88.43% as a mean average precision (mAP) for 15 multiple objects.

Automotive paint defect detection plays a crucial role in the automotive production process. Current research on visual defect detection methods is mainly based on supervised learning, which requires a large number of labeled image samples for model training. The labeling work is not only time consuming but also expensive, seriously hindering the testing and application of these models in practice. To address this issue, this study proposes a new method for automotive paint defect detection based on a semi-supervised training strategy. First, a semi-supervised automotive paint defect detection framework, which can use labeled and unlabeled samples to reduce the cost of data labeling effectively, is presented. Then, a spatial pyramid pooling fast external attention module that introduces an external attention mechanism is proposed to improve the traditional YOLOv7 network structure, called YOLOv7-EA, to obtain good detection performance. This network acts as a detector to generate high-quality pseudo labels for the unlabeled samples, providing additional data to train the model; meanwhile, it performs the final detection task. Lastly, a Wise-intersection over union loss function that considers the quality of the anchor box is introduced to reduce the interference of low-quality samples and improve the convergence speed and detection accuracy of the model. Using this method, we can accomplish the task of automotive paint defect detection with a small number of labeled image samples. Experimental results on the automotive paint defect dataset show that mean average precision (mAp)@.5, mAp@.75, and mAp@.5:.95 are superior to other methods under the condition of 10% and 15% labeled data, achieving good defect detection performance.

This study aims to determine the optimal batch size in training the YOLOv8 model for object detection in autonomous vehicles. With the increasing need for accurate and efficient object detection technology, this study explores the effect of batch size variation on the performance of the YOLOv8 model. The dataset used in this study is a traffic simulation dataset from CARLA, obtained from the Roboflow universe, consisting of 1719 images divided into training, validation, and testing data. The research methodology includes data collection, pre-processing, and data analysis using the YOLOv8 technique with different hyperparameter settings. The results showed that increasing the number of epochs and batch size contributed to the increase in the mean Average Precision (mAP) value of the model. The best training scheme was identified with the highest mAP value of 98.2%, using 100 epochs, batch size 32, and image resolution 640x640. These findings provide important insights for further development in object detection technology, as well as provide guidance for researchers who want to optimize training parameters for object detection models using YOLOv8 in the context of autonomous vehicles. This research is expected to serve as a reference for future studies in this field.Kata kunci: YOLOv8, object detection, autonomous vehicle, optimal batch size, CARLA dataset, mean Average Precision (mAP), hyperparameters, model training

An improved faster R-CNN algorithm for assisted detection of lung nodules

Since the glass outer screen of a cell phone is the main sensory part of the human eye when using a cell phone, the advantages and disadvantages of the cell phone screen directly affect people's sense of use. Therefore, the defect detection requirements for cell phone screens are high and need to meet the needs of high-volume factory inspection. Most of the traditional defect detection methods use visual methods, the detection results are overly dependent on the subjectivity and experience of workers, the efficiency of this method is low, and the accuracy is poor. Currently, machine learning-based detection methods are applied in numerous industries. In this paper, a faster Regional Convolutional Neural Network (R-CNN) with multi-head attention mechanism for defect detection of cell phone screen is proposed. To enhance the network's capability in extracting feature information, a four-head attention mechanism is added to the last convolutional layer of the ResNet50 network. An improved Region of Interest (ROI) Align is proposed to replace the original ROI Pooling to reduce the localization error of cell phone screen defects. Replace the original Rectified Linear Unit (ReLU) activation function with the Copy Exponential Linear Unit (CELU) activation function to expedite the convergence capability of the network. Finally, by comparing with other classical model training, the evaluation results indicate that the proposed method achieved an average accuracy of 95.71%, which is a 5.34% improvement compared to the original faster R-CNN network.

Object detection is an important part of remote sensing image analysis. With the development of the earth observation technology and convolutional neural network, remote sensing image object detection technology based on deep learning has received more and more attention and research. At present, many excellent object detection algorithms have been proposed and applied in the field of remote sensing. In this paper, the object detection algorithms of remote sensing image is systematically summarized, the main contents include the traditional remote sensing image object detection method and the method based on deep learning, emphasis on summarize the remote sensing image object detection algorithm based on deep learning and its development course, then we introduced the rule of performance evaluation of object detection and datasets that commonly used. Finally, the future development trend is analyzed and prospected. It is hoped that this summary and analysis can provide some reference for future research on object detection technology in remote sensing field.

ABSTRACTThe detection of defects on industrial surfaces is essential for guaranteeing the quality and safety of products. Deep learning‐based object detection methods have demonstrated impressive efficacy in industrial applications in recent years. However, due to the complex and variable shape of defects, the similarity between defects and background, large intra‐class differences, and small inter‐class differences lead to low classification accuracy, it is a great challenge to achieve accurate defect detection. To overcome these challenges, this research proposed a novel network specifically designed for defect detection. First, a feature extraction network, ResDCA‐Net, is constructed based on deformable convolution and lightweight multi‐scale attention, where deformable convolution can adaptively adjust to extract features of defects with complex and variable shapes. Second, the lightweight multi‐scale attention module is constructed, which uses multi‐branch and cross‐space fusion to obtain the complete feature space attention map, thereby improving the defect feature attention and reducing the background feature attention. Third, to enhance the classification and localization accuracy, an attention‐based decoupled prediction module is proposed to ensure that the classification and regression branches of the model can focus on their required features. Finally, extensive comparative experiments indicate that the proposed approach performs best, achieving 83.7% and 83.4% mean Average Precision (mAP) on the GC10‐DET and NEU‐DET datasets, respectively. The effectiveness of the proposed individual modules is further validated in ablation experiments, which demonstrate the excellent performance and potential in defect detection tasks.

Methods that enable the visual inspection of solar panels are currently in demand, as a huge number of solar panels are now being deployed as a sustainable energy source. One of the solutions for inspection automation is an end-toend deep learning framework, but this is not recommended for this problem because such a framework requires not only powerful computational resources, but also a large-scale classbalanced dataset. In this study, we present a cost-effective solar panel defect detection method. We emphasize the spatial feature of defects by utilizing an attention map that is generated by a pre-trained attention mechanism that can give attention on stroke ends, gathering, and bends. We define and extract 13 statistical features from the attention map, and then feed them into conventional machine learning model. Therefore, we no longer require energy depleting models such as end-to-end neural classifiers to discriminate between non-defective and defective panels. Five conventional machine learning models and one stateof-the-art (SOTA) deep learning model—i. e., EfficientNet—are used to generalize the experimental results. The results of the comparative experiments indicate that our approach, which includes attention mechanism recycling and statistical feature extraction, is guaranteed to provide cost-effective defect detection in general with performance that is competitive with that of recent SOTA. In future research, we expect that our approach can be adopted in other defect detection tasks such as steel or film manufacturing processes.

Batched-image detection model and deployment method for tunnel lining defects using line-scan cameras based on experiments study

Object detection has been known as the core of computer vision and attracted much research attention in recent years especially because of its close relationship with video analysis and image understanding. According to the abundant research on object detection, many traditional object detection methods have been proposed. This paper introduces some famous traditional methods, which are based on SIFT, HOG, SURF, and ORB. However, due to the characteristics of large amount of computation and simple training structure, the traditional detection method has low detection speed. With the fast rise of deeper learning, stronger devices are implemented to address the problems that exist in conventional architectures. In the architecture of the network, training and optimization functions etc., these models are special. In this paper, we review the frameworks for object detection based on deep learning. We begin our review with the methods based on Convolutional Neural Networks. Then typical methods of object detection and some helpful modification to improve detection performance are introduced. Moreover, the methods based on YOLO and SSD are introduced. In fact, despite of the same basis of algorithm or features, the performance and features of different methods are various. Thus in this paper we analyze the features and architecture of each method. This also offers some research to equate various approaches and draw some concrete conclusions. In this paper, we also introduce some typical datasets used for testing or training the object detection model. This paper made a systematic classification and summary in the object detection field, which can be meaningful and useful for the scholars who started to learn about it.