Simultaneous Determination of Multi Rice Quality Parameters Using Image Analysis Method

Abstract

Determination of rice quality is a key function for rice research and rice industry. Traditionally, rice quality parameters were performed separately by human visual inspection. It is labor-intensive and time-consuming. In this paper, a novel analysis method was proposed to determine multi rice quality parameters simultaneously. On the glass of a flatbed scanner, 35 g rice kernels was spread and imaged. Then, the length, width, aspect ratio, head rice yield, percentage of chalky rice, chalkiness, and transparency grade were obtained after simply processing the acquired image. The developed method for the determination of rice quality was tested on 507 milled rice samples and compared to the traditional manual analysis method. The regression coefficients (R) were 0.9916, 0.9691, 0.9938, 0.9929, 0.9649, and 0.9377 for length, width, aspect ratio, head rice yield, percentage of chalky rice, and chalkiness, respectively. The accuracies of length, width, aspect ratio, and head rice yield determined by the developed method were 100, 100, 100, and 92.1 % respectively. The results indicated that the developed method was a useful and alternative method for determining rice quality.