The apple images were hard to be identified at a faster speed and a higher accuracy because of fuzzy and uncertain factors existing in the color image boundary pixels, so in order to overcome the disadvantages above, a model combined quantum genetic algorithm and fuzzy neural network was built up which showed the capability of global search capability and adaptation. In the proposed model, quantum genetic algorithm was used to optimize the initial value of adjustable parameter in fuzzy neural network, which avoided redundant iteration and the incline to fall into the local minimum value of traditional BP algorithm. The experimental results showed that the proposed model achieved accuracy of 100% for the uneven color samples, 96.86% for sunlight influenced samples, 94.29% for the adjacent samples, and 92.31% for the overlapping samples.