Interactive Insect Pest Recognition Method on Mobile Devices

摘要

In the field of agricultural research, it is difficult to achieve satisfactory results in the analysis of crop pests. The results are even worse especially on the images obtained from field. It's hard to calculate the number of pests on a blade accurately, due to the effects of illumination and complex background. In order to obtain more reliable results and make the method have more value in practical applications, the need for precise positioning of the crop blades and reducing the impact of ambient light of the image are very urgent for further pests analysis. Targeted at pests analysis on mobile devices, this paper presents a segmentation algorithm and a de-illumination algorithm for crop insect pest automatic counting, and implements them on mobile devices. The experimental results show that the proposed method has good recognition performance with high efficiency on the images acquired in the field and greenhouse.