Share Email Print
cover

Proceedings Paper

Hyperspectral visible-near infrared imaging for the detection of waxed rice
Author(s): Mantong Zhao
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Presently, unscrupulous traders in the market use the industrial wax to wax the rice. The industrial wax is a particularly hazardous substance. Visible-near infrared hyperspectral images (400-1,000 nm) can be used for the detection of the waxed rice and the non-waxed rice. This study was carried out to find effective testing methods based on the visible-near infrared imaging spectrometry to detect whether the rice was waxed or not. An imaging spectroscopy system was assembled to acquire hyperspectral images from 80 grains of waxed rice and 80 grains of non-waxed rice over visible and near infrared spectral region. Spectra of 100 grains of rice were analyzed by principal component analysis (PCA) to extract the information of hyperspectral images. PCA provides an effective compressed representation of the spectral signal of each pixel in the spectral domain. We used PCA to acquire the effective wavelengths from the spectra. Based on the effective wavelengths, the predict models were set up by using partial least squares (PLS) analysis and linear discriminant analysis (LDA). Also, compared with the PLS of 80% for the waxed rice and 86.7% for the non-waxed rice detection rate, LDA gives 93.3% and 96.7% detection rate. The results demonstrated that the LDA could detect the waxed rice better, while illustrating the hyperspectral imaging technique with the visible–near infrared region could be a reliable method for the waxed rice detection.

Paper Details

Date Published: 13 November 2014
PDF: 13 pages
Proc. SPIE 9276, Optical Metrology and Inspection for Industrial Applications III, 927622 (13 November 2014); doi: 10.1117/12.2073990
Show Author Affiliations
Mantong Zhao, Univ. of Shanghai for Science and Technology (China)


Published in SPIE Proceedings Vol. 9276:
Optical Metrology and Inspection for Industrial Applications III
Sen Han; Toru Yoshizawa; Song Zhang, Editor(s)

© SPIE. Terms of Use
Back to Top