An odor recognition system for apple freshness detection is proposed for rapid evaluation of apple quality. The system consists of a gas sensor array connected to a data acquisition system and a computer with pattern recognition software. By collecting and detecting the aromas emitted by apples

stable system response values are used to set up cluster analysis and classification models. The successive projections algorithm（ SPA） is employed to optimize the sensors array

and the results proved that the optimized sensor array is feasible to distinguish freshness of apple. Back propagation（ BP） neural network classification model is built. Accuracy of testing reach 98. 67% based on the optimized sensor array

and this study can accurately distinguish apples with different freshness. It can be used as fast evaluation method for apple quality.