Yak Body Size Measurement and Weight Estimation Based On 3D Point Cloud

In order to reduce the limitations of obtaining yak body size parameters by traditional two-dimensional images, such as measurement parameters limitations, measurement errors, the experimental method based on 3D point cloud and support vector machine SVM is used to obtain yak body size and weight information quickly and accurately. The yak side point cloud is collected based on the depth camera RealSense D435i. With the help of point cloud preprocessing and geometric features, the yak body size measurement points are extracted, and the yak body size parameters are calculated and the weight estimation is completed. The results show that, compared with the results of manual measurement, the mean absolute percentage error of yak body height is 2.06%, the mean absolute percentage error of body oblique length is 1.71%, and the mean absolute percentage error of chest circumference is 4.22%. The overall error is small. The yak weight estimation model based on support vector machine SVM is studied, and the mean absolute percentage error of the model is 4.6%, which proves the feasibility of this research method. It shows that the method of yak body size measurement and weight estimation based on 3D point cloud can meet the requirements of contactless automatic measurement of yak, save time and labor costs, and make Qinghai yak industry more intelligent and professional.