Abstract:  Fruit disease is one of the important factors affecting the healthy of fruit trees, fruit quality and yield. Timely and accurate detection of fruit disease information and precise application control are of great significance for preventing the occurrence and prevalence of major diseases in orchards and ensuring stable and excellent fruit yield. With the demand of largescale, intelligent and highefficiency development of modern agriculture, visual and spectral detection technology has gradually developed into one of the important technologies for the detection of fruit diseases, because of its advantages of nondestructive testing, largescale and highefficiency. This paper reviews the research progress of computer vision and spectral technology in the field of fruit disease detection, and introduces that the image processing technology has good explanatory ability, which is conducive to the combination with plant protection agronomy research. Deep learning technology has better precision and generalization. Transmission spectrum technology can be used to detect internal diseases of fruit. Reflectance spectroscopy can be used to detect the surface diseases of fruits and leaves, and achieve classification. Finally, the future optimization and application of computer vision and spectral detection technology are summarized, and the practical production and application prospects of fruit disease detection are prospected, so as to provide reference for the research of fruit disease detection.

Key words: fruit diseases, disease detection, computer vision, spectral technology

摘要： 水果病害是影响果树健康生长、果实品质和产量的重要因素之一，及时、精准地掌握果树的病害信息并进行精准施药管控，对防范果园重大病害的发生和流行，保障水果的稳产优产具有重要意义。随着现代农业朝规模化、智能化和高效率的发展需求，视觉和光谱检测技术因具有无损检测、可规模化和高效率等优点，逐渐发展为检测水果病害的重要技术之一。梳理国内外机器视觉和光谱技术在水果病害检测应用领域的研究进展，介绍图像处理技术有较好的解释性，有利于与植保农艺研究相结合；深度学习技术有较好的精度和泛化性；透射光谱技术可用于检测果实内部病害；反射光谱技术可用于检测果实、叶片表面病害，并实现分级。最后，总结未来机器视觉与光谱检测技术优化和应用的方向，并展望水果病害检测的实际生产应用前景，以期为水果病害检测研究提供参考与借鉴。

关键词: 水果病害, 病害识别, 机器视觉, 光谱技术