Variety Identification of Single Rice Seed Using Hyperspectral Imaging Combined with Convolutional Neural Network

The requirements of reliability, expeditiousness, accuracy, consistency, and simplicity for quality assessment of food products encouraged the development of non-destructive technologies to meet the demands of consumers to obtain superior food qualities. Hyperspectral imaging is one of the most promising techniques currently investigated for quality evaluation purposes in numerous sorts of applications. The main advantage of the hyperspectral imaging system is its aptitude to incorporate both spectroscopy and imaging techniques not only to make a direct assessment of different components simultaneously but also to locate the spatial distribution of such components in the tested products. Associated with multivariate analysis protocols, hyperspectral imaging shows a convinced attitude to be dominated in food authentication and analysis in future. The marvellous potential of the hyperspectral imaging technique as a non-destructive tool has driven the development of more sophisticated hyperspectral imaging systems in food applications. The aim of this review is to give detailed outlines about the theory and principles of hyperspectral imaging and to focus primarily on its applications in the field of quality evaluation of agro-food products as well as its future applicability in modern food industries and research.

Hyperspectral imaging in the visible and near infrared (VIS-NIR) region was used to develop a novel method for discriminating different varieties of commodity maize seeds. Firstly, hyperspectral images of 330 samples of six varieties of maize seeds were acquired using a hyperspectral imaging system in the 380–1,030 nm wavelength range. Secondly, principal component analysis (PCA) and kernel principal component analysis (KPCA) were used to explore the internal structure of the spectral data. Thirdly, three optimal wavelengths (523, 579 and 863 nm) were selected by implementing PCA directly on each image. Then four textural variables including contrast, homogeneity, energy and correlation were extracted from gray level co-occurrence matrix (GLCM) of each monochromatic image based on the optimal wavelengths. Finally, several models for maize seeds identification were established by least squares-support vector machine (LS-SVM) and back propagation neural network (BPNN) using four different combinations of principal components (PCs), kernel principal components (KPCs) and textural features as input variables, respectively. The recognition accuracy achieved in the PCA-GLCM-LS-SVM model (98.89%) was the most satisfactory one. We conclude that hyperspectral imaging combined with texture analysis can be implemented for fast classification of different varieties of maize seeds.