摘要
起球图像的分割是织物起球客观评定方法研究的一个重要步骤,同时也是一个非常困难的问题。针对这种图像的分割,提出了一种基于脉冲耦合神经网络(PCNN)的分割方法,该方法是利用织物表面毛球区域与未起球区域灰度强度的不同,根据PCNN神经元是否点火来分割毛球图像。通过对5种不同起球等级图像进行的分割实验证明,算法是有效的。
Pilling image segmentation is not only an important stage on objective evaluation of fabric pilling but also a difficult problem. A new method of fabric pilling image segmentation based on pulse-couple neutral networks(PCNN) is put forward in the paper. According to different gray intensity between the field of pilling and the field of unpilling, feature of fabric pilling are segmented for PCNN cell firing. Five pilling grade test image are segmented with this algorithm, and the validity of this method is proved by experimental results.
出处
《中国图象图形学报》
CSCD
北大核心
2007年第7期1230-1233,共4页
Journal of Image and Graphics
基金
陕西省科技厅国际合作资助项目(2004WK-06)
陕西省教育厅产业化重点项目(05JC13)
关键词
脉冲耦合神经网络
织物起球
图像分割
算法
pulse couple neutral networks (PCNN), fabric pilling, image segmentation, algorithm