An apple classification algorithm based on wavelet and fuzzy is proposed in order to get the edge of apple image accurately and realize apple automatic classification. Firstly, the apple image is transformed by omnidirectional wavelet, and then processed by fuzzy algorithm. The edge of the apple image is extracted by adaptive threshold, and then the area of the apple image is obtained by using the diffuse filling algorithm. Finally, according to the characteristics of the apple quasi circle, the area is transformed into the straight diameter, and the apple classification is completed according to the diameter. The simulation results show that the accuracy of the algorithm is more than 98% for three grades of apples, which shows that the algorithm can be used in the classification detection of apples.