针对光纤收发PCB板的传统手工检测与校正中存在的检测速度慢、检测效率低和校正效率低等问题,课题组提出一种基于HALCON的检测与校正方案。通过图像降噪、截取模板和模板匹配等处理手段快速定位目标位置并计算标准位置;通过比较识别位置与标准位置的坐标差,进而判断贴片位置是否合格,对于不合格产品可由电机平台重复推动与检测直至产品合格。实验结果表明:新系统能够快速、准确、有效地检测出不合格品;检测准确率达99. 4%,平均检测时间约为367 ms,校正后合格率为100%。新系统的检测精度、检测速度以及校正后产品合格率均符合工业生产的要求,能够应用于工业生产。
To solve the problems of low detection speed,low detection efficiency and low correction efficiency in traditional manual detection and correction of optical fiber transceiver PCB,a detection and correction scheme based on HALCON was proposed. The target position was quickly located and the standard position was calculated by image denoising,template interception and template matching. The coordinate deviation between the identified position and the standard position was compared to determine whether the patch position was qualified. For the unqualified product,the motor platform could be repeatedly pushed and tested until the product was qualified. The experimental results show that the system can detect the unqualified products quickly,accurately and effectively. The detection accuracy is 99. 4%,the average detection time is about 367 ms,and the qualified rate after correction is 100%. The detection accuracy and speed of the system and the qualified rate of the products after correction meet the requirements of industrial production and can be applied in industrial production.
Light Industry Machinery