针对贴片机生产中手机屏蔽罩视觉定位问题,分析了现有模板匹配算法的不足,提出了采用轮廓特征边提取算法实现屏蔽罩的视觉定位;算法针对屏蔽罩图像特点首先应用二值化及Blob点分析方法去除图像中的干扰点,再采用Laplacian变换提取出边缘轮廓,然后由外边缘轮廓提取出线段集并用屏蔽罩尺寸信息过滤出特征边,最后由特征边计算出屏蔽罩的中心及偏转角度;实验效果表明该算法的识别速度和精度都能满足实际生产的需要,并具有较强的鲁棒性。
Aimed at the image locating problem with shielding ease, this paper analysis the shortage of general pattern match algorithm then present a fast locating algorithm based on characteristic edges detect of shielding case. First, it uses binarization and blob analysis method to remove the small particles; Next, it performs a Laplacian transform to get the edge image; And then finds the line segment sets and filter with the size information of the ease to get the characteristic edges; Finally deduces the center and rotate angle with the characteristic edges. The result shows the locating speed and the accuracy of the algorithm is well and meets the practical application. The algorithm also has strong robust.