Pest Identification System based on YOLOv5

Abstract

Forest pests and diseases are a global problem. The key to control forestry diseases is to accurately identify the species and severity of pests. How to use artificial intelligence and image recognition technology to detect forestry pests is an important challenge and opportunity. This paper presents a new method for forestry pest identification based on YOLOv5 algorithm. In addition, in order to unify the system and expand the flexibility of the future system, we adopted the B/S/S structure to develop the pest identification system. The system uses the camera to shoot images and transmits the data to the background recognition. The experimental results show that our system can detect the target pests more accurately and conveniently, which is helpful for the actual prevention and control of forestry pests.