【摘要】 针对机器视觉中锂电池极柱难以准确识别的问题，研究一种基于视觉的锂电池极柱位置识别方法。首先对图像进行预处理以提取锂电池区域和位姿；然后通过布尔运算得到锂电池的首末端图像；最后对方块形电池与圆柱形电池分别比较矩形度和轮廓线段总长度，快速、准确识别出极柱位置。实验表明，该算法识别准确率达98%，可满足锂电池自动拆解的需要。更多还原

【Abstract】 Aiming at the problem that the poles of lithium batteries are difficult to accurately identify in machine vision, this study researches a vision-based method for identifying the position of lithium battery poles. First,image preprocessing is performed to extract the lithium battery region and its pose. Then, the images of the battery’s positive and negative ends are obtained through Boolean operations. Finally, for prismatic and cylindrical batteries, the rectangularity and the total length of contour segments are compared respectively to quickly and accurately identify the pole positions. Experiments show that the recognition accuracy of this algorithm reaches98%, meeting the requirements for automated disassembly of lithium batteries.更多还原