Study on identification of rice seeds by chemical oscillation fingerprints

Abstract

A variety of nonlinear chemical oscillation fingerprints were studied which were produced by the interaction between rice seeds and a B–Z oscillation system. The chemical constituents and their relative content in the rice seeds could be indirectly reflected by the oscillation fingerprints, and the characteristic parameters of the oscillation fingerprints could be automatically extracted by a program. The results showed that all rice seed samples used in the experiment could be identified accurately by both principal component analysis and cluster analysis, and the identification process could be efficiently completed in 10 to 20 minutes. The method has many advantages in terms of application, such as simple operation, low cost, rapid identification and so on.