Automated calculation of heart girth measurement in pigs using body surface point clouds
2019
Wang, Ke | Zhu, Dehai | Guo, Hao | Ma, Qin | Su, Wei | Su, Yang
As increasing number of studies for body measurement purposes in both the animal breeding and production stages, several papers have already evaluated the feasibility of consumer depth sensors for linear body measurement of livestock including pigs and cows. However, taking heart girth measurements is one of the major problems in livestock farming and the current studies remain under-evaluated for automated extraction of livestock girth measurements. In this paper, we develop a novel automatic measurement method for pig heart girth from body surface point clouds. Firstly, we preprocess the original point cloud data with methods including segmentation and pose normalization. Secondly, we resort to point cloud cross-section feature to detect the measuring position of the heart girth. Thirdly, the heart girth point clouds is achieved by slicing point cloud on the measuring position. Finally, the length of the heart girth is obtained by approximating the perimeter of the curve fitted by the heart girth point clouds. Experimental results show that the correct detection rate of measuring position of pig heart girth is 96.5% and the average relative error for heart girth from our method is 7.87%, which demonstrates the efficacy of the proposed measurement method. The program of automated calculation of heart girth measurement in pigs using body surface point clouds and its corresponding data set can be downloaded freely from https://github.com/LiveStockShapeAnalysis/Girth-fitting-methods to livestock industry and research.
显示更多 [+] 显示较少 [-]