Fast Detection Method of Rice Storage Quality Based on Machine Olfaction

Abstract

A fast detection method of rice storage quality was studied in this paper. The smell quality detection of stored rice was as a research object, and the odor data was collected by the gas sensor array. The method of principal component analysis was used for feature dimension reduction. Finally the BP network and stochastic resonance algorithm were used for classification respectively. The test results showed that the correct recognition rate was 93.33% by using the classification method of stochastic resonance. The machine olfaction system proposed in this paper can achieve the purpose of testing the rice storage quality non-destructively, rapidly and accurately, which provides a new way for food quality nondestructive testing research.