Body measurements of dairy cows using a structure from motion (SfM) photogrammetry approach

The possibility of frequently monitoring calf body growth allows early recognition of health anomalies, and accordingly can reduce occurrence of problems connected to infertility or other diseases. In the last fifty years, manual measurement has been the most straight forward way to evaluate the animal body parameters. However, this approach is a laborious operation and may be stressful both for the cows and the stockman. In the present work, Structure from Motion (SfM) is studied and proposed as a low-cost and non-invasive technique for three-dimensional reconstruction of cow body. Analyses carried out on a fiberglass model clearly show how SfM has the potential to provide a high resolution three-dimensional reconstruction, which can be applied for quantitative extraction of body parameters. The approach has been implemented to get rid of cow movements, based on a segmented approach, where different portion of the body have been imaged during different frames acquisition tasks, focused firstly on the top and secondly on the sides. Experimental tests demonstrated how body portions ranging between 20 and 40% of the whole cow area could be properly reconstructed, thus allowing an adequate data set for extraction of body indices. Keywords: body measurement, calf growth, low-cost sensor, 3D modeling, structure from motion.