论文部分内容阅读
文章介绍了基于FFT特征匹配在计算机视觉检测芯片管脚间距中的具体应用的研究 ,在对芯片管脚间距进行检测时 ,将CCD所拍摄到的芯片图像转换为灰度图像 ,之后提取边缘特征 ,再利用芯片的一个标准管脚的边缘特征图像作为模板 ,经图像匹配 ,阈值检验后 ,得到与管脚数目相同的极大值点 ,各极值点之间的距离即得各管脚间距。通过此方法可以快速测出管脚间距 ,并判断是否合格 ,对于相近领域内的其它定量分析测量有一定的参考价值。
This paper introduces the research on the application of FFT feature matching in the pin pitch of the computer vision inspection chip. When the chip pitch is detected, the chip image captured by the CCD is converted into a gray image, and then the edge feature is extracted , And then use the edge of a standard pin chip image as a template, the image matching, threshold test to get the same number of pins with the maximum number of points, the distance between the extreme points that have each pin pitch . This method can quickly detect the pin spacing, and determine whether the qualified, for other quantitative analysis of similar areas have a certain reference value.