A Research Review of Pest Identification and Detection Based on Deep Learning

Efficient and accurate identification of pests will effectively help solve agricultural pest problems, promote the intelligence, professional and greening of crop pest control. Traditional pest identification and control methods are inefficient, and the use of artificial intelligence technologies such as deep learning and computer vision can significantly reduce human costs and largely improve the efficiency and accuracy of the pest identification process. A review is conducted on deep learning-based pest identification methods, introducing the application and development of different deep learning networks, analyzing the latest research progress of domestic and international scholars in the field, and comparing the performance and characteristics with each other. Existing pest recognition datasets are summarized to provide a reference for in-depth research on the application of pest recognition and detection.