垩白米粒的计算机图像识别

Discernment of Chalky Rice Kernel with Image Processing System

摘要: 自行开发研制了计算机图像处理系统，用于优质稻谷国家标准GB/T 17891-1999中质量指标垩白度及垩白粒率的检测。利用该系统对6种粳米和2种籼米进行测定，结果表明，该方法具有客观性、准确性、快速性和可重复性等特点，在收购优质稻米进行快速分等定级中具有良好的应用前景。

Abstract: An image processing system was developed which was specially used for the detection of the quality indexes of chalkiness and percentage of chalky rice kernels listed in the national standard GB/T 17891-1999. Six round-grained and two long-grained nonglutinous varieties of rice were inspected with the aid of the system. Its characteristics were well objective, accurate, quick and repeatable. The system is a potential solution for rapid quality determination in rice purchasing practice.