Not stated in source 2025 Needs review

Taking the labour out of labour detection – a computer vision approach

Visit original source

Abstract

Difficulty during parturition has negative consequences for both animal welfare and farm profitability on cattle farms. Early intervention has been shown to be a key requirement for minimising losses caused by calving problems, however, this requires high levels of monitoring to ensure that problems are promptly detected and resolved. Current methods of detecting calving either rely on labour intensive manual observations or cow-based sensors which are often expensive and require the placement of physical sensors on each animal. In this study, we explore the potential of computer vision technology to accurately predict calving in dairy cattle and highlighted the advantages of both object and pose detection approaches in comparison to current methods.

Keywords

Not available

Specialisations

Similar theses