小波变换在零件边缘检测中的应用

目前对机械零件的尺寸、口径等参数的测量大多采用接触测量,而利用零件图像的边缘信息进行非接触测量,则是一种新的更好的检测方式。用此方法可以对机械零件图像的边缘进行精确的检测,再根据相应的数学关系得到相关参数的值,不但可以提高零件检测的精度,还可以大大节省人力和物力。模拟边缘检测技术的具体过程,依次采用最佳阈值迭代算法、小波变换模极大值法、边缘连接局部处理法对零件图像边缘信息进行边缘检测和提取,并对检测过程进行了重点分述。最后,通过一个具体实验,验证了小波技术与其他技术结合使用所获得的良好效果,从而进一步显示了小波变换在边缘检测技术中应用的巨大潜力。

Application of Wavelet Transform in Mechanic Edge Detection

At present,parameters of the mechanic accessories,for example,size and caliber,are still measured by contact measure.Non-contact measure is a new method of better effect by utilizing the image edges of the mechanic accessories. The image edges of mechanic accessories can be detected precisely via this method,then correlative parameters will be able to be obtained by mathematic expressions.This kind of measurement not only can improve the accuracy of the accessories detection, but can save the manpower and resources greatly.According to the concrete processes of edge detection,several algorithms were used in turn in this paper.These algorithms were iterative algorithm of the best threshold,the maximum module algorithm of wavelet and the local processing algorithm for edge jointing,and at the same time,detecting procedures were introduced briefly. Eventually,a concrete experiment was put forward to examine the results.The results demonstrated that the methods for detecting the edges were successful by adopting the wavelet transform and other algorithms.Thereby,the tremendous latent capacities of wavelet transform were further revealed in the area of edges detection.