The application of Deep Learning can significantly improve the accuracy of plant disease detection and recognition, and overcome the limitations of traditional methods. This article first collects and introduces some publicly available datasets of plant disease images, and then provides a systematic review of the research and the applications of Deep Learning in this field in recent years, including early detection and recognition algorithms and the development of deep learning-based algorithms. This article further points out that management should be carried out through the entire lifecycle of plant growth. In each stage of plant growth, advanced image acquisition and recognition technologies such as UAV technology and video image processing should be used to conveniently, promptly, and massively obtain the multimodel data of plant growth status in large areas. Combined with Deep Learning technology, plant disease detection should be widely applied to better serve agricultural modernization and automation, which is also one of the future development directions. Finally, this article provides valuable reference for the in-depth research and development of plant disease identification.