CAJ | 학술논문

对电子制造行业中电控板质量的视觉自动检测技术进行研究。通过分析电控板的存在特征，选用Harris算子进行电控板图像特征点的提取，然后进行待检测图像与模板图像的特征点匹配，并设置约束条件去除误匹配点，再判断特征点像素坐标是否在对应的元器件的像素区域内，实现漏件的检测；同时利用图像的灰度值模板匹配法，结合改进的序贯相似性检测法，实现电控板错件的检测。试验结果表明该算法对电控板错件和漏件的检测正确率达96.8%，检测的算法运行时间在2.862s内，具有一定的精确性和实时性，为制作企业生产线的信息智能化提供技术支持。
대전자제조행업중전공판질량적시각자동검측기술진행연구。통과분석전공판적존재특정，선용Harris산자진행전공판도상특정점적제취，연후진행대검측도상여모판도상적특정점필배，병설치약속조건거제오필배점，재판단특정점상소좌표시부재대응적원기건적상소구역내，실현루건적검측；동시이용도상적회도치모판필배법，결합개진적서관상사성검측법，실현전공판착건적검측。시험결과표명해산법대전공판착건화루건적검측정학솔체96.8%，검측적산법운행시간재2.862s내，구유일정적정학성화실시성，위제작기업생산선적신식지능화제공기술지지。
Studies visual automatic detection technology of electric control board quality in the electronics manufacturing industry. Through analyz-ing the existing characteristic of electric control board, chooses the Harris operator to extract the image feature points of electric control board, carries on the feature points matching between be detected and the template images, and sets up the constraint conditions to re-move false matching points, at last judges the feature points pixel coordinates whether in the pixel area of corresponding components, real-izes the detection of missing part; meanwhile uses the template matching method of image grey value and the improved sequential simi-larity method to realize the detection of electric control board wrong parts. The test results show that the testing accuracy for wrong and missing parts of electric control board is 96.8%, and the detection algorithm running time within 2.862s, which has a certain accuracy and real-time performance, and provides technical support to the information intelligence of production enterprise production line.