Abstract: This paper takes the fruit sorting control process as the research object, and establishes the sorting control algorithm based on RGB image detection method. In the process, the asynchronous image acquisition mode is used to acquire fruit image, and the median filter and Gaussian filter are used to remove fruit image noise. The global automatic threshold segmentation method is used to extract fruit image features, so as to realize the recognition and classification of fruit color features and surface area features. The feature data is compared with the set feature threshold In order to realize the identification of fruit grade, the robot is used for fruit sorting. The experimental results show that the fruit feature data obtained by RGB image detection method and manual measurement method are highly consistent.