一种面向密封条长度检测的模板匹配优化算法

针对现有的汽车密封条长度测量设备中使用模板匹配算法检测特征点位置时,密封条自身旋转导致匹配效果差的问题,提出一种基于Hough变换的自适应模板更新算法。该算法在传统模板匹配方法的基础上,通过设定阈值对模板匹配效果进行判断;对于匹配效果不好的图像进行Hough变换,结合特征点的几何特征进行识别;通过特征点的位置与尺寸信息,将其转化为新的模板用于下一次的模板匹配,实现模板的自适应更新。试验结果表明,该算法弥补了传统模板匹配法对图像旋转异常敏感的问题,提高了识别的稳定性,同时保证了检测效率满足现场的要求。

Template Matching Optimization Algorithm for Seal Length Detection

Aiming at the problem of poor matching effect caused by the rotation of the sealing strip itself when the template matching algorithm is used in the existing automobile sealing strip length measurement equipment to detect the position of the feature point,an adaptive template update algorithm based on Hough transform is proposed.Based on the traditional template matching method,the algorithm judges the template matching effect by setting a threshold.Hough transform is performed on the images with poor matching effect,combined with the geometric characteristics of the feature points for recognition.Through the location and size information of the feature points,it is transformed into a new template for the next template matching,and the adaptive update of the template is realized.The test results show that the algorithm compensates for the abnormal sensitivity of the traditional template matching method to image rotation,improves the stability of recognition,and at the same time ensures that the detection efficiency meets the requirements of the scene.