Research on the Classification and Recognition Method of Bovine Milk Somatic Cells Based on Gaussian Mixed Model

Mastitis is a common disease in dairy cows, and early diagnosis is crucial for improving cow health management and milk quality. Somatic cell count (SCC) is a key indicator of mastitis, but traditional methods struggle to effectively classify different types of somatic cells. This paper proposes a somatic cell classification method based on the Gaussian Mixture Model (GMM), which initializes clustering through the K-means algorithm, improves the covariance matrix of the Gaussian distributions, and iteratively optimizes the E-step and M-step of the Expectation-Maximization (EM) algorithm for accurate classification of different somatic cell types. Experimental results show that the improved GMM model can accurately classify the somatic cells of dairy cows, with an average classification accuracy of 97.95%. Compared to traditional methods, this approach offers significant advantages in classification accuracy and reliability, providing a new technical solution for early diagnosis of mastitis and milk quality control.