Fecha de publicación: 14/06/2024 Fuente: WIPO "processing tomato" The invention provides a tomato maturity grading method based on vision and touch fusion, which comprises the following steps: acquiring tomato image information, and adopting two color space conversion schemes to obtain color feature information of the surface of a tomato; meanwhile, the maximum between-cluster variance method is used for image segmentation, so that the image processing operand is effectively reduced; the image difference is further increased by adopting a gamma transformation algorithm, and the surface glossiness feature information of the tomatoes is obtained; a touch sensor and a PID control method are adopted to obtain surface roughness feature information, the surface roughness feature information is fused with image feature information, the feature information is input into a BP neural network for training, a tomato maturity grading model is obtained, namely the corresponding relation between the tomato feature information and the tomato maturity grade, and the tomato maturity grading model is obtained for tomatoes meeting the maturity requirement. And outputting a detection grading result. According to the method, automatic, efficient and accurate maturity grading of tomatoes is realized, and meanwhile, the model has good robustness.