基于视觉及概率统计的谷粒在线计数系统设计

摘要

针对目前粮食谷粒人工计数存在的检测效率低、检测时间长的问题,设计了基于LabVIEW与Matlab谷粒在线快速计数系统.通过设计的谷粒采集装置采集谷粒同时获取谷粒图片,并采用距离变换与分水岭相结合的分割算法进行图像分割,同时利用概率密度估计理论统计图像中谷粒区域概率最大的像素个数,从而得出该幅图像中的谷粒个数.该系统谷粒计数准确率可达97％,提高了传统谷粒计数的速率以及准确度,可安装在联合收割机上并为后续监测谷粒的破碎率、含杂率提供技术支撑.%To address the shortages of inefficiency and time-consuming in manual counting for grains causes of human subjectivity,a high-speed counting system for grain developed by LabVIEW and Matlab is designed.The grain is collected by the grain collection device,and the grain image is segmented by using the segmentation algorithm that combined the distance transformation with the watershed.At the same time,the probability density estimation theory is used to calculate the maximum number of pixels in the grain region.Thereby the number of grains in the image is obtained.The test results show that the recognition accuracy of the system is 97％,which improves the efficiency and accuracy of the traditional system and can provide the technical support to monitor the crushing rate and impurity rate of grains in the combine harvester tank.