Sim K S, Kho Y Y, Tso C P, Nia M E, Ting H Y
Faculty of Engineering & Technology, Multimedia University, Melaka, Malaysia.
Scanning. 2013 Mar-Apr;35(2):75-87. doi: 10.1002/sca.21037. Epub 2012 Jul 9.
Detection of cracks from stainless steel pipe images is done using contrast stretching technique. The technique is based on an image filter technique through mathematical morphology that can expose the cracks. The cracks are highlighted and noise removal is done efficiently while still retaining the edges. An automated crack detection system with a camera platform has been successfully implemented. We compare crack extraction in terms of quality measures with those of Otsu's threshold technique and the another technique (Iyer and Sinha, 2005). The algorithm shown is able to achieve good results and perform better than these other techniques.
利用对比度拉伸技术对不锈钢管道图像中的裂纹进行检测。该技术基于一种通过数学形态学的图像滤波技术,能够暴露裂纹。裂纹被突出显示,同时有效地去除噪声,且仍保留边缘。一个带有摄像头平台的自动裂纹检测系统已成功实现。我们将在质量度量方面的裂纹提取结果与大津阈值技术和另一种技术(Iyer和Sinha,2005年)的结果进行比较。所示算法能够取得良好的效果,并且比其他技术表现更好。