【摘要】 传统播种均匀性变异系数的测量需人工定位和计数，耗时耗力，效率较低。为提高播种均匀性变异系数测定的速度和精度，研究利用实时点云重建技术实现播种均匀性变异系数的自动测量。首先通过深度相机获取种子图像和环境稠密点云信息；接着进行图像分割，计算种子形心的图像坐标；然后再从实时点云信息中筛选出种子的三维坐标；最后将该三维坐标转换至初始坐标系，并向水平面投影，进而根据投影结果实现播种均匀性变异系数的测定。试验结果表明，所提出的方法与人工测量相比，在水平方向的平均定位误差分别为2.08 mm和2.443 mm，单次测量耗时小于0.5 s，播种均匀性变异系数误差为0.4%。更多还原

【Abstract】 The measurement of the coefficient of variation of traditional sowing uniformity requires manual positioning and counting, which is time-consuming, labor-intensive, and inefficient. In order to improve the speed and accuracy of measuring the coefficient of variation of sowing uniformity, the automatic measurement of the coefficient of variation of sowing uniformity using real-time point cloud reconstruction technology was studied. Firstly, the seed images and environmental dense point cloud information are obtained through a depth camera. Then, image segmentation is performed to calculate the image coordinates of the seed centroid. Then, the three-dimensional coordinates of the seed are filtered out from the real-time point cloud information. Finally, they are transformed into the initial coordinate system and projected onto a water plane. Based on the projection results, the measurement of the coefficient of variation of sowing uniformity is achieved. The experimental results show that compared with manual measurement, the proposed method has average positioning errors of 2. 08 mm and 2. 443 mm in the horizontal direction, with a single measurement time of less than 0. 5 s, and an error of 0. 4% in the coefficient of variation of sowing uniformity.更多还原