Differentiating pre- and post-grazing pasture heights using a 3D camera: a prospective approach

Grasslands management involves the monitoring of both animal and plant components. Recent precision livestock farming developments have focused on high-rate monitoring of grazing animals to enhance livestock productivity and welfare. The evolution of grass resource during the grazing process is not being overlooked by graziers and researchers, but grass characteristics, such as height, dry matter content, productivity or density, are still measured using low frequency and sometimes destructive and time-consuming methods; such as quadrat, sward-sticks, rising plate meters. This study investigated the potential of using 3D cameras to assess sward physical characteristics. Main objectives were: (1) to define the correct way to capture images, particularly the camera position above the ground and, (2) to assess if differences in sward height were detectable. Couples of images differing in grass height were captured on the same spot with a 3D camera at different above-ground heights (30, 40, 50 cm) on a ryegrass-white clover pasture. Pregrazing height was 15cm and post-grazing sward was simulated by cutting at 2 cm. Histograms of intensity performed on greyscale images showed differences between preand post-grazing sward. As expected, overall darker pixels were observed for pre-grazing images (p<0.01) and whiter pixels for post-grazing images (p<0.01), indicating longer distances consistent with lower forage biomass. Images taken at a distance of 30 and 40 cm could identify these differences. Further developments require improving the calibration of the camera and developing image analysis method to estimate more plant characteristics such as density or dry matter content. 238  Precision Livestock Farming '17