摘要
为实现基于脱绒棉种外观特性对红种、破碎棉种的分选,搭建了脱绒棉种外观质量检测的硬件装置,并基于Visual C++6.0开发了检测的软件系统,提取了适合脱绒棉种分级的3个关键的特征参数:颜色、圆形度、粒型;基于BP神经网络实现了脱绒棉种的分级,通过试验验证,该神经网络的拟合性较好。
The paper is to achieve the detectation to the red and crashed seeds, build up the hardware device of delinted cotton seeds＇ appearance quality, develope the software system of detectation based on Visual C ＋ ＋ 6.0, distill three key character parameters suited to delint cotton seeds grading： color, circular degree, seed type; attain the delinted cotton seeds grading based on BP NN. Through experiment it testified that the emulation of this BP NN is preferable.
出处
《石河子大学学报（自然科学版）》
CAS
2006年第6期761-764,共4页
Journal of Shihezi University(Natural Science)
基金
教育部科学技术研究项目(206168)
关键词
脱绒棉种
外观质量
机器视觉
检测
delinted cotton seeds
appearance quality
machine vision
detectation
