摘要
利用先进的基于CCD的裂缝测试仪采集到试件的表面裂缝。为了计算裂缝宽度,必须准确识别出裂缝边缘。针对裂缝特征,提出了直方图法、结合法、梯度法和改进Roberts法等4种图像测量分析方法,并分别从原理上进行了解释,同时设计了实验标定图进行图像处理和测量,给出了四种分析方法的处理效果图、裂缝宽度及其误差分析。分析结果表明,类判别和数学形态学的结合法在测量裂缝方面效果较好,并将其应用于实际测量中,实践证明了该方法的有效性。
The crack image of sample is captured by the CCD of the advanced crack test instrument. In order to calculate the crack width, the edges of crack must be recognized accurately. Four image-measurement methods, which are the histogram method, the combination method, the grads method and the betterment Robertgs method, are proposed based on the surface crack feature and interpreted by utilizing the principles of image recognition. At the same time, the calibration images designed are experimented by four image methods. The crack width and error analysis are presented by comparing their testing results. The combination method of category distinguishing and mathematics morphology is applied in the practice measurements. The result shows that it has good effect on processing images of crack.
出处
《光学与光电技术》
2008年第5期45-48,共4页
Optics & Optoelectronic Technology
基金
国家自然科学基金(10472026)资助项目
南京工程学院科研基金(KXJ08016)资助项目
关键词
裂缝测试仪
图像测量
类判别
数学形态学
crack test instrument
image measurement
category distinguishing
mathematics morphology
