A Mathematical Morphology Edge Detection for Weld Image

摘要：

In order to improve the sharpness of the detected edge by the conventional morphology edge detection, a modified mathematical morphology detection is proposed for weld edge detection. In the proposed method, basic gray-level morphological operations dilation and erosion are modified to restrain edge expansion from the boundary to the flat intensity area. The proposed method weakens the response of noises and strengths the response of edge points. The experiments on the weld images show that the proposed method performs good on suppressing noises, and improves accuracy on weld edge detection.

关键词： mathematical morphology edge detection weld image

作者: Dawei Sun Yan Cai Yixiong Wu

作者单位: School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, China

会议类型: 国际会议

会议名称: 2010 International Conference on Circuit and Signal Processing(2010年电路与信号处理国际会议 ICCSP 2010)

会议地点: 上海

会议语种:英文

页码: 193-196

在线出版日期: 2010-12-25（万方平台首次上网日期，不代表论文的发表时间）