Leroy, T., Laboratory for Agricultural Buildings Research, Catholic University Leuven, Kasteelpark Arenberg 30, B-3001 Heverlee, Belgium
Aerts, J.-M., Laboratory for Agricultural Buildings Research, Catholic University Leuven, Kasteelpark Arenberg 30, B-3001 Heverlee, Belgium
Eeman, J., Laboratory for Agricultural Buildings Research, Catholic University Leuven, Kasteelpark Arenberg 30, B-3001 Heverlee, Belgium
Maltz, E., Institute of Agricultural Engineering, Agricultural Research Organization, Volcani Center, PO Box 6, 50520 Bet Dagan, Israel
Stojanovski, G., Institute of Agricultural Engineering, Agricultural Research Organization, Volcani Center, PO Box 6, 50520 Bet Dagan, Israel
Berckmans, D., Laboratory for Agricultural Buildings Research, Catholic University Leuven, Kasteelpark Arenberg 30, B-3001 Heverlee, Belgium

A camera-based method was developed to determine the body condition score (BCS) value of individual cows in an automatic way. Seventy five back-view images of dairy cows (Holstein) were used in developing the algorithm for automatic body condition scoring. The BCS of each new test cow was predicted by comparing the contour of the cow with the contours of the set of reference cows, whose BCS was evaluated by experts. The results are promising because the error of the prediction of the BCS with the camera-based method is on average of the same order of magnitude as the error of the human experts (0.27 vs. 0.25). In this way, automatic monitoring of the BCS of dairy cows, as a part of a precision livestock farming system, might become realistic in practice in the future.