摘要
爆腰米粒的检测是米的品质检测中的重要内容之一。本文阐述了运用计算机图象处理技术检测爆腰米粒的的三种方法——直方图傅里哀系数判别法、直方图波峰检出法和灰度突变判别法,三种方法的识别功能、检测精度和简便程度各不相同。通过对八个品种798个样本的实测,表明用计算机视觉代替感官检测米的爆腰是可行的。
Testing of cracked particles of rice is one of major contents in rice quality testing. By applying the technique of computer image processing, three methods for testing the cracked rice particles-histogram Fourier coefficient decision method, histogram peak detection method and gray-eruptive decision method, which are different in recognition function, testing precision and symplicity, are described. The testing result of 798 samples of 8 varieties shows that testing of cracked with computer visual perception instead of sense organ is feasible.
出处
《农业工程学报》
EI
CAS
CSCD
北大核心
1992年第3期104-112,共9页
Transactions of the Chinese Society of Agricultural Engineering
基金
国家自然科学基金
关键词
计算机
图象处理
米
品质
检测
Computer image processing, Rice quality testing, Cracked rice.
