自适应带钢表面缺陷图像边缘增强方法

中国机械工程 ›› 2010, Vol. 21 ›› Issue (05): 571-574.

自适应带钢表面缺陷图像边缘增强方法

魏玉兰;颜枫;颜云辉;李兵

东北大学，沈阳，110004

出版日期:2010-03-10 发布日期:2010-03-22
基金资助:
国家自然科学基金资助项目(50574019)；国家863高技术研究发展计划资助项目(2008AA04Z135)
National Natural Science Foundation of China（No. 50574019）;
National High-tech R&D Program of China (863 Program) （No. 2008AA04Z135）

Adaptive Edge Enhancement Method for Steel Strip Surface Defect Image

Wei Yulan;Yan Feng;Yan Yunhui;Li Bing

Northeastern University，Shenyang，110004

Online:2010-03-10 Published:2010-03-22
Supported by:

National Natural Science Foundation of China（No. 50574019）;
National High-tech R&D Program of China (863 Program) （No. 2008AA04Z135）

摘要/Abstract

摘要：

针对带钢表面缺陷图像去噪后边缘模糊的问题，提出了一种基于邻域信息的自适应双倒数滤波方法。该方法结合邻域内像素的几何位置信息与灰度信息进行滤波，通过引入与边缘有关的梯度算子自适应调整几何位置信息中的距离，在图像灰度变化小的背景区域取较大的权值，增强图像的平滑效果，以达到去噪的目的；在图像灰度变化大的边缘处取较小的权值，以达到边缘增强的目的。实验结果分析表明，该方法去噪和边缘保持效果好。

关键词:

带钢, 灰度相似度, 几何邻近度, 边缘增强, 双倒数滤波

Abstract:

According to neighborhood information an adaptive double-reciprocal filter was proposed to improve the edges which were degraded by denoising. In this method, geometric vicinity and photometric similarity were combined to filtering. In order to improve the effect, an edge related gradient coefficient was also chosen, which adaptively adjusted the original distance variance and brought a big weight in the background to remove the noises and enhanced the smoothness. On the contrary, a small weight was brought in the edge area. The blurriness of the edge can be obviously avoided by the proposed method. The experimental results show that this method is effective to eliminate the noises, and the edge detection is significantly improved.

Key words: steel strip, photometric similarity, geometric closeness, edge enhancement, double-reciprocal filter

中图分类号:

TP391.41

魏玉兰, 颜枫, 颜云辉, 李兵.

自适应带钢表面缺陷图像边缘增强方法

[J]. 中国机械工程, 2010, 21(05): 571-574.

WEI Yu-Lan, YA Feng, YA Yun-Hui, LI Bing.

Adaptive Edge Enhancement Method for Steel Strip Surface Defect Image

[J]. China Mechanical Engineering, 2010, 21(05): 571-574.

[1]	丛家慧, 颜云辉. 视觉注意机制在带钢表面缺陷检测中的应用 [J]. 中国机械工程, 2011, 22(10): 1189-1192,1221.
[2]	魏玉兰, 颜云辉, 李兵, 张尧, 李骏. 邻域灰度与空间特征相结合的互信息配准方法研究 [J]. 中国机械工程, 2011, 22(4): 439-443.
[3]	夏瑞雪, 卢荣胜, 刘宁, 董敬涛. 基于圆点阵列靶标的特征点坐标自动提取方法 [J]. J4, 201016, 21(16): 1906-1910.
[4]	夏瑞雪, 卢荣胜, 刘宁, 董敬涛. 基于圆点阵列靶标的特征点坐标自动提取方法 [J]. 中国机械工程, 2010, 21(16): 1906-1910.
[5]	王永慧, 颜云辉, 赵大兴, 赵久梁. 版图分类法在带钢表面缺陷识别中的应用研究 [J]. 中国机械工程, 2010, 21(05): 562-566.