Pest Detection in Crop Images Based on OTSU Algorithm and Deep Convolutional Neural Network

With the development of computer image detection technology in agriculture, the accurate detection of crop pests under complex background has become an important issue in agriculture. Due to various forms and complex environments, some pest images cannot be accurately detected by existing detection algorithms. In order to improve the detection accuracy, a deep convolutional neural network based on feature fusion is proposed. This algorithm is based on Mask R-CNN network and OSTU, introduces automatic threshold segmentation algorithm. In the feature extraction stage, an improved threshold segmentation algorithm is introduced, and then the feature data generated by segmentation is used to replace the original feature maps. The experiment on crop pest detection shows that this detection algorithm proposed in this paper can effectively detect crop pests and achieve the effect of identification and instance segmentation.