Prediction possibility of milk yield from udder measurements using digital image analysis on holstein cows

Ozkaya S.

INDIAN JOURNAL OF ANIMAL RESEARCH, vol.49, no.3, pp.388-391, 2015 (SCI-Expanded) identifier identifier


A photographic survey of 43 multiparous lactating cows was carried out by recording the images between morning and evening milking time (around noon time) using digital camera from posterior angle. Images were then processed using Image Tools V3.00 software to determine Udder measurements (UMs) of cows from the images in centimetres. The linear parameters recorded were rear udder depth (UD), rear udder width (11W) and rear udder area (UA). Regression analysis of milk yield on each of the independent variables was performed. Regression coefficients, for all UMs, taken together were 65.5%. The single regression coefficients were23.0, 0.2 and 61.0for UD, UW and UA, respectively. The correlation coefficients between milk yield and UD, UW and UA were 0.48 (P>0.05), 0.04 (P>0.05) and 0.78 (P<0.05), respectively. The results showed that, the milk yieldof cows could be predicted using digital image, using udder area as a parameter. However, the prediction possibility was higher when all UMs were used.