THE POTENTIAL OF THE PROCESSING OF DIGITAL IMAGES IN THE DIAGNOSIS OF SUBCLINICAL MASTITIS IN MILK COWS
somatic cell count, lactoperoxidase, mammary gland health, RGB system
The objective of this work was to develop a simple tool for image processing, focusing on the diagnosis of subclinical mastitis quickly, effectively and at low cost. The study was carried out in the municipality of São Gonçalo do Amarante (RN), where milk samples were collected from Jersey, Gir and Guzerá cows. Somatic Cell Counting (CCS) and quantitative and qualitative enzymatic identification analyzes were performed, the samples were photographed with a smartphone and the CCS and lactoperoxidase (LP) parameters were quantified by the Red-Green-Blue (RGB) color system . The results of the PCA (Principal Component Analysis) model showed that a classification principle due to the presence of subclinical mastitis in dairy cows, providing a variance of 98.4%. The intensities of the LP and CCS RGB channels increased from right to left as LP concentration increased, especially channel B. A supervised method was trained using PLS-DA (Partial Least Squares Discriminant Analysis), this presented to the group test a sensitivity of 96% and a specificity of 81.6% of the results obtained. Image processing by the RGB color system presents possible innovation to be used as an analytical method for CCS and LP in bovine milk.