摘要
木棉和棉同属于纤维素纤维,化学性质非常相似,采用传统方法不能有效地鉴别.通过光学显微镜获得的纤维纵向图像发现,木棉纤维纵向较为光滑,没有类似棉的天然转曲,因此可以利用图像处理的方法进行纤维识别.首先提出了一种纤维分离算法,通过轮廓和骨架的分离,从而有效地分离交叉纤维.然后利用灰度分布特征和经典模式识别方法实现了木棉和棉的自动识别.
Kapok and cotton both belong to the cellulose fiber and their chemical properties are very similar,so using traditional methods can not recognize them efficetively. In fiber longitudinal images acquired through light microscope, kapok fiber＇s longitudinal appearance is smooth, while cotton fiber has natural twist, so the digital image processing method can be used to recognize kapok and cotton. A new algorithm is proposed, which can separate crossing fiber effectively through the separation of the contour and skeleton. Then, gray level distribution and classic pattern recognition method are used to realize the automatic recognition of cotton and kapok.
出处
《东华大学学报（自然科学版）》
CAS
CSCD
北大核心
2017年第1期95-101,133,共8页
Journal of Donghua University(Natural Science)
基金
陕西省教育厅科学研究计划资助项目(14JS039)
关键词
木棉
棉
数字图像处理
交叉纤维
纤维分离
自动识别
kapok
cotton
digital image processing
crossing fiber
fibers＇ separation
automatic recognition
