DAIRY COW HOOF MOVEMENT ANALYSIS BY VISION TECHNIQUES Claudia Bahr1, Astrid Koppenol1, Arno Pluk1, Xiangyu Song1, Willem Maertens2, Annelies Van Nuffel2, Juergen Vangeyte2, Bart Sonck2, Daniel Berckmans1 1 2 Division of Measure, Model & Manage Bioresponses (M3-BIORES), Katholieke Universiteit Leuven, Kasteelpark Arenberg 30 - B3001 Leuven, Belgium E-mail: [email protected] Tel: +32 16 321 726 Fax: +32 16 321 480 Institute for Agricultural and Fisheries Research Technology & Food Unit - Agricultural Engineering, B-9820 Merelbeke, Belgium Early detection of dairy cow lameness is important for effective treatment and ailment prevention. Most of the lameness cases are caused by infectious and non-infectious diseases in hoof region. The objective of this research is the analysis of hoof movement by using vision techniques to detect lameness in dairy cows. First experiments were done on a commercial farm with 108 milking cows when they passed a corridor from the milking parlour to the feeding area. A Guppy F036C camera with a SV-03514 lens was located 1,5 m away from the corridor and 30 cm high above the ground. The locomotion of cows was scored visually within a scale of 1 (sound), 2 (moderately lame) and 3 (severe lame). The focus was on the parameter “ankle angle. It was calculated two times for a step of each hoof: first in the image taken just before the hoof touches the floor (touch angle) and secondly in the image taken just after the hoof is released from the floor (release angle). Table 1. Means of touch and release angle in a group of 63 cows All cows Cows score 1 Cows score 2 Cows score 3 Mean touch angle in degrees FL FR HL HR 62,1 60,6 58,7 58,3 61,9 60,5 58,9 58,8 62,9 63,4 56,6 59,1 61,9 59,8 54,5 56,6 Mean release angle in degrees FL FR HL HR 122,9 120,1 107,8 105,8 123,6 119,7 108,0 105,5 125,5 122,2 106,8 104,6 121,9 122,0 105,5 102,7 Table 1 gives an overview about the mean of touch angle and release angle in numbers for each hoof. In case of occuring lameness, a decrease of touch and release angle is expected. Regarding this expectation the results show a trend of decrease in touch and release angle within an increasing gait score especially in the hind hoofs.