Wood defect detection based on YOLOV5

In the process of wood processing, improving the detection accuracy of wood defects is of great significance to improve the safety and utilization rate of wood use. At present, convolutional neural network is mainly used to detect wood, but it has shortcomings such as inaccurate detection of small targets, weak ability to capture global feature information and low detection efficiency, which leads to the occurrence of wrong detection and missing detection in the detection process. In response to the above problems. In this paper, based on the YOLOV5 network structure, a small target detection network module is introduced to solve the problem of inaccurate detection of small targets, ensure that fine-grained information is not lost, and make the feature fusion mode simple and effective. Then, a high-precision detection network module is introduced to realize the combination of convolutional neural network structure and transformer structure to solve the problem of weak ability of the original network to capture global feature information.In this paper, CBAM attention mechanism is introduced to reduce and suppress the influence of similar interference features on network detection, so that the network can distinguish the defect features correctly. In addition, the lightweight network structure of YOLOV5S guarantees real-time monitoring performance. The experimental results show that compared with the original network, the average detection accuracy of the improved algorithm is increased by 6.623%, which can effectively improve the detection accuracy of network structure and improve the detection ability of small target defects.