Research on Machine Vision Based Inspection of Rice Appearance Quality
By employing the developed inspection system using an embedded computer, the inspection algorithms for rice appearance quality are discussed in this study. The algorithms can detect several appearance features of rice, such as the chalky grains ratio, degree of chalkiness, yellow-colored grains, grain shape, and so on. A novel watershed algorithm based on prior knowledge is used for separating conjoint grains. A BP neural network is chosen for detecting chalky grains; hue, for detecting yellow-colored grains; and an algorithm in the polar coordinate system, for calculating the long and short axis of each region to detect the grain shape. The experimental results reveal that the detection accuracies of the chalky grains ratio, chalkiness degree, yellow-colored grains, and grain shape are 98％, 99％, 90％, and 94％, respectively.
Ming Sun Shuhuai Zhang Dong An Yaoguang Wei
College of Information and Electrical Engineering China Agricultural University Beijing P.R. China 1 Faculty of Agriculture and Life Science Hirosaki University Hirosaki Japan 036-8561
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）