Possibility of Early Detection of Bovine Mastitis in Dairy Cows Using Thermal Images Processing

摘要

Bovine mastitis (BM) is a prevalent condition on dairy farms, affecting both livestock health and reducing profitability. This study investigated the feasibility of diagnosing BM in Holstein dairy cattle using thermography. To increase the detection between healthy cattle and unhealthy one and to better compare the results from thermal images, a number of parameters including somatic cell count (SCC) and California mastitis test (CMT) were adopted. The result of non-parametric Spearman's Rho test showed that there was aninverse correlation (R=-0.97) between SSC factor and milk production records. Bovine Mastitis diagnosis results obtained from processing thermal images showed that the average temperature difference between unhealthy and healthy tissues was 0.44 ?C. The detection accuracy of this method was 57.3%. The results from processing the thermal images showed that the thermal imaging camera was able to detect small temperature difference on the skin sometimes due to the effects of factors, such as non-uniform light irradiation and the presence or absence of hair or skin lesions on udder surface, which are not necessarily the studied factors. Therefore, it is suggested that parameters such as udder’s touching to groin, hair on udder, reduced light effect and skin lesions in order to improve the precision in the thermographic diagnosis of BM also be controlled future studies.