图像处理在零件检测中的应用

Application of image procession in part detection

介绍了一种依据图像处理技术检测机械零件的测量方法。以面阵CCD为图像传感器,通过图像采集卡将机械零件的二维图像输入到计算机中。在对原始输入图像进行直方图校正和边缘保持滤波处理后,对得到的较为平滑的零件图像进行边缘检测,提取目标曲线,并据此计算出零件的各参数值。最后通过MATLAB进行仿真,论证了该方法的高效准确,实用的优点。此种测量方法非常适合于微小的接触测量难以准确测量的机械零件的参数检测,具有广阔的应用前景。

This paper introduces a new method for measuring machine element which is based on image processing. With paneled CCD as the image sensor, the two-dimensional pictures of machine part is inputted to a computer through an image sensor. After the revising of square diagram and filtering of the edge keeping, edge detection is carried on to the smoother picture to get target curve, and each parameter value of the parts is figured out on these grounds. Through simulating with Matlab, the method is proved effi- cient, accurate and practical. This method is suitable for the permanent detection of the element which is small and hard to measure accurately through contact, and have vast of applied foreground.