Automatic lameness detection based on 3D-video recordings
T. van Hertem, Ephraim Maltz, Aharon Antler, Victor Alchanatis, A.A. Schlageter Tello, C. Lokhorst, C.E.B. Romanini, Stefano Viazzi, Claudia Bahr, Daniël Berckmans, I. Halachmi · Socio-Environmental Systems Modeling · 2013
Manual locomotion scoring for lameness detection is a time-consuming and subjective procedure. Therefore, the objective of this study is to quantify the classification performance of a computer vision based algorithm for automated lameness scoring. Cow gait recordings were made during four consecutive night-time milking sessions in an Israeli dairy farm with a 3D-camera. A live on-the-spot assessed 5-point locomotion score was the reference for the automatic lameness score evaluation. A dataset of 1436 cows with automatic lameness scores and live locomotion scores was used for calculating classification performance. The analysis of the automatic scores as independent observations led to a correct classification rate of 50.4% on a 5-point level scale. When allowing a 1 unit error on the 5-point level scale, a correct classification rate of 87.6% was obtained. The obtained tolerant binary correct classification rate was 88.6%. The automated lameness detection system obtained a tolerant correct classification rate of 88.6%.