Chili Pepper Pests Recognition Based on HSV Color Space and Convolutional Neural Networks

With the expansion of the planting area, the planting of chili peppers faces an increasing number of pests and diseases, and the development of computer vision recognition has become a crucial means of preventing and controlling these pests. Currently, the majority of crop disease and pest recognition algorithms use the RGB color space, which makes the algorithm model more dependent on the green channel's feature information. This paper utilizes the HSV color space to process chili pepper images, allowing convolutional neural networks (CNNs) to extract extra features from limited image samples. In comparison to the RGB color space, the proposed algorithm has superior precision and recall, as well as superior generalization and application capabilities. Therefore, it has broad application scenarios in the field of computer identification of crop pests and diseases and can be used extensively.