Related Experiment Video
Updated: Jan 7, 2026
Dual Raster-Scanning Photoacoustic Small-Animal Imager for Vascular Visualization
Published on: July 15, 2020
Design and Verification of a Non-Contact Body Dimension Measurement System for Jiangquan Black Pigs Based on
Zhao Ma1, Shiyin Li1, Zhanchi Ren1
1Shandong Provincial Key Laboratory for Livestock Germplasm Innovation & Utilization, College of Animal Science and Technology, Shandong Agricultural University, Taian 271018, China.
None:
To address inefficiencies, pig stress from traditional manual body dimension measurement, and environmental interference in existing automated technologies, this study designed and validated a non-contact measurement system for Jiangquan black pigs based on dual-view (top + side) depth vision (Intel RealSense D455). Key dimensions (body length/width/height, chest depth) were accurately extracted via depth map calibration, dynamic scaling, and U-Net segmentation. Chest girth was estimated using the Ramanujan ellipse perimeter model (MAE = 4.15 cm, R2 = 0.908) and integrated as the core parameter for body weight prediction in an empirical formula. This experimental dataset comprises 30 pigs sourced from a single farm, with body weights falling within a limited range (30-100 kg). All dimensions achieved R2 > 0.9, with top-view body width performing best (R2 = 0.9424, MAE = 1.9 cm). Body weight prediction yielded R2 = 0.957 and MAE = 5.1 kg. The system completes measurements in 24 ± 4 s with low hardware costs and stress-free operation, making it suitable for precision breeding in small-to-medium pig farms.
