基于凸包的电池片边缘缺陷检测算法

SOLAR CELL EDGE DEFECT DETECTION ALGORITHM BASED ON CONVEX HULL

摘要:

为了实现电池片分选机的快速分拣,需要加快电池片图像各项检测项目的速度。其中,电池片边缘缺损与破片的检测是把控产品质量的首要检测项目。已有的检测方法图像处理时间较长,处理过程可复用性差,处理结果受环境影响大。因此,提出一种通过凸包判断电池片边缘是否存在残缺,同时计算缺陷位置像素深度的快速边缘缺损检测办法。该方法对成像环境要求低,且可以提供电池片像素中心及偏转角度等电池片姿态信息,缩短了其他检测过程的时间。最后,实验证明该算法的平均时间消耗小于50ms,同时,大量电池片检测结果的误检率和漏检率均小于0.5%。

Abstract:

This In order to realize the rapid sorting of solar cells slice sorter, it is necessary to speed up each image process detection items. Among them, the detection of edge defect and fragment of solar cells is the most important item to control product quality. The existing detection methods have a long processing time, poor reusability of the processing process, and the processing results are greatly affected by the environment. Therefore, this paper proposed a fast edge defect detection method based on convex hull to determine whether the edge of the battery is defective or not, and to calculate the pixels depth of the defect location. This method has low requirement for the imaging environment, and can provide the attitude information of the cell at the same time, such as the center of the cell pixel and the deflection angle. These results Shorten the time of other detection processes. Finally, experiments show that the time consumption of the algorithm is less than 50 ms, and the false detection rate and missed detection rate of a large number of battery test results are less than 0.5%.specification is set for the theses to be published in Computer Applications and Software, including fonts, margins, page size and print area.