针对废有色金属分选中物料动态位置计算的问题,提出了一种基于二值图像形心计算的改进算法,以保证物料位置计算的准确性和高效性.对基于X射线透射技术的废有色金属分选进行了分析,利用图像处理技术对探测数据进行了处理.考虑到待分选有色金属废料形状、尺寸、姿态各异,移动速度快的特点,选用了高精度的连通区标记法,并在二值图像形心算法的基础上进行了改进,实现了连通区标记后一次遍历计算多个物料目标形心,解决了同一物料的探测数据被分割至两幅图的问题,完成了物料位置计算.实验结果表明,该计算方法能快速、准确地计算有色金属物料的动态位置,算法的准确率可达92.3％,平均每幅图耗时30.3 ms.%Aiming at the problem of calculation of materials' dynamic position in scrap non-ferrous metal sorting,an improved algorithm based on the shape center algorithm of binary image was proposed to ensure the accuracy and efficiency of position calculation.The scrap non-ferrous metal sorting based on X-ray transmission technology was analyzed,image processing technology was used to process the detection data.Considering non-ferrous metals materials had the characteristics such as differences in shape,size and attitude,high speed,high precision connected components labeling was adopted and an improved method was proposed based on the shape center algorithm of binary image.The shape centers of multiple objects could be calculated by scanning the image for once when the connected components labeling was completed.The problem of the detection data of the same material was divided into two images can also be solved and the position of non-ferrous metals materials could be calculated eventually.The results of experiment show that the algorithm can calculate the position of the non-ferrous metals materials quickly and accurately,the accuracy of the algorithm can reach 92.3％ and the average running time is 30.3ms per image.
展开▼
