Aiming at the shortcomings of traditional PCB board inspection methods, such as low detection efficiency and low detection accuracy, a PCB board defect detection method with improved YOLOv5 model was proposed. In order to improve the precision of small target defect detection, the BiFPN based network connection method is constructed, which makes full use of the feature information of different scales. In order to better capture the position of target defects, we introduced Coordinate Attention mechanism to make model positioning and target capture more accurate. The experimental results show that compared with the original YOLOv5 model, the mean average accuracy of the proposed method for detecting PCB surface defects is improved by 3.2%.