Abstract: In view of the drawbacks of apple grade identification in China, which still relies on human sense organs, ths paper presents a processing method on the basis of the technology of computer vision and digital image. The method of grade identification can change the competitiveness of the parameters by changing the weights of parameters in input layer of the Learning Vector Quantization(LVQ). The improved LVQ neural network algorithm was applied in the process to identify the grade of apples, which was proved effective by experiment, reaching the correct identification rate of 88.9% with good stability.