Development of an automatic rice-quality inspection system

The need has arisen in rice-drying facilities in Japan for an automatic method to measure quality aspects of rice when it arrives at the drying facility. A near-infrared (NIR) transmission instrument was used to obtain NIR spectra of damp rough rice and damp brown rice. Calibration models were developed from the original spectra and reference analysis data to determine moisture and protein content of the samples. A visible light (VIS) segregator was used to determine sound whole kernel of brown rice. The precision and accuracy of the NIR instrument and the VIS segregator were found to be sufficiently high to determine moisture and protein content, and sound whole kernel ratio. An automatic rice-quality inspection system was consequently developed. The system consisted of a rice huller, a rice cleaner, an NIR instrument and a VIS segregator, and it was controlled by a computer. Based on the rice-quality information, this system enabled rough rice transported to a rice-drying facility to be classified into six qualitative grades. (C) 2003 Elsevier Science B.V. All rights reserved.