摘要
通过研究油茶果籽和皮的颜色特性,将线阵工业相机采集的油茶果籽和皮图像分割成若干正方形像素块,基于模糊聚类原理,根据特征降维、一致性度量和模式识别的相关理论,研究油茶果的颜色识别分类方法,实现一种基于彩色线阵相机的动态模糊聚类色选算法,并将算法应用于通道式单面色选机的预选和实时分选程序中。应用表明,算法对新鲜的油茶果籽分选选净率达到98%。
The sorting of the camellia seeds and shells mainly relies on manual selection currently.The color characteristics of camellia oleifera seeds and shells were investigated by taking camellia fruit image through industrial linear CCD camera.The picture was divided into several square blocks of pixels,based on the principle of fuzzy clustering,the color sorting method of camellia oleifera was researched according to the feature dimension reduction,consistency feature selection and pattern recognition theories.A dynamic fuzzy clustering color selection algorithm based on color linear CCD camera was realized,and it was applied to the primary and real-time separation process of the channel-type color sorter.The recognition rate of fresh camellia oleifera seed could reach 98%,and therefore the algorithm has a high promotional value.
出处
《中国农机化学报》
2015年第4期94-99,共6页
Journal of Chinese Agricultural Mechanization
基金
湖北省自然基金项目(CDB085)
关键词
油茶果
模糊聚类
图像处理
色选
camellia oleifera
fuzzy clustering
image processing
color sorting
