Abstract: The image features of tea diseases are complex, diverse, and fuzzy, posing a threat to the health and population of tea.At present, there is a lack of corresponding disease degree evaluation system, and the weight of evaluation indicators cannot accurately reflect the current growth status of tea.OpenCV is a powerful image processing tool, and its image edge defect recognition algorithm can reduce the misdiagnosis rate of diseases and has good background noise processing effects.This study used OpenCV to train and analyze the proportion of tea lesions, and obtained qualitative evaluation results.At the same time, the importance of the four characteristic indicators, namely the size of tea lesions spot size, tea texture, tea color, and tea size was quantitatively evaluated by experts.The membership degree theory of fuzzy mathematics was introduced for comprehensive judgment, and a comprehensive expression for tea disease was established.Based on the appropriate evaluation weights, further determine the impact of different factors on the degree of tea disease.The practical application results indicated that the combination of OpenCV and fuzzy mathematics evaluation methods, through reliable and stable image recognition technology, can more scientifically and reasonably evaluate the degree of tea diseases, providing reference value for the healthy growth and quality assurance of tea.