基于机器视觉的嵌入式珍珠分拣器研究与制作

Research on the Embedded Pearl Sorte Based on Machine vision

本珍珠分拣器通过摄像头采集珍珠图像,经过单片机处理后输出分拣控制命令使得珍珠分拣装置自动按照设定的分类标准对珍珠进行分类,并将分类的结果数据保存。由像素点对应实际几何面积的比例计算得到珍珠的大小;从图像中珍珠的有效像素的RGB值中得到珍珠的颜色;计算珍珠边界重心代替圆心计算,用遍历法找到目标的最大和最小半径,以此判断珍珠的圆度。珍珠分拣装置根据单片机传送的分拣命令把珍珠按照设定的标准分类。在最后的实验样机实物测试中,文章设计的珍珠分拣器基本实现把混杂的珍珠样品按照设定的标准进行分类,并把分拣的结果送到上位机监控软件显示。

This sorting machine collecting images by camera and then the MCU will send a massage to put pearl in rightclassification according to the criterion that we set, and save the result of classification data at last. People can calculate thesize of the pearl according the number of pixels and the color of the pearl by image pixels RGB value and get theroundness of pearl from the maximum and minimum radius of pearl. After the image processing, the pearl sorting devicewill classify pearls according to the standards set before.In final test, the result shows the pearl sorting device can classify themixed pearls according to the standards we set and send the sorting result to PC monitoring software.