基于改进区域生长法的羊体点云分割及体尺参数测量

针对传统羊体尺测量中测量耗时、应激大的问题,采用主成分分析、随机采样一致性算法和改进的区域生长法,基于三维点云对羊体尺参数测量进行研究。结果表明:1)使用主成分分析和随机采样一致性算法能计算羊体点云的法向量和曲率;2)改进的区域生长法能准确地分割出羊体区域,并且避免了外点的干扰;3)在羊体点云数据上选取体尺测点,计算羊体长、体高、臀高、胸深体尺参数,并与实测值比较,4种体尺参数的最大相对误差为2.36%,测量精度较高。试验证明改进的区域生长法能准确地对羊体点云进行分割,依据选取的体尺测点,能够实现羊体尺参数的无接触测量。

Point cloud segmentation and measurement of the body size parameters of sheep based on the improved region growing method

Aiming at the problem of time-consuming and strong stress reaction in traditional sheep size measurement, principal component analysis, random sampling consistency algorithm and an improved region growing method are used to study the measurement of sheep body size parameters based on 3D point cloud. The results show that: 1)The normal vector and curvature of sheep point cloud can be calculated by principal component analysis and random sampling consistency algorithm; 2)The improved region growing method can accurately segment sheep body region and avoid the interference of outliers; 3)The measurement points of sheep body size are selected from the point cloud data to calculate the parameters of body length, body height, hip height and chest depth, and the results are compared with the measured values. The maximum relative error of the four body size parameters is 2. 36%, and the measurement accuracy is relatively high. The experiment proves that the improved region growing method can segment sheep point cloud accurately, and the non-contact measurement of sheep body size parameters can be realized by selected measuring points.