Detection of Surra Disease in Microscopic Images of Horse Blood Using the K-Nearest Neighbor (KNN) Method

Abstract

Surra's disease is an infectious disease caused by the parasite Trypanosoma evansi and attacks farm animals such as horses, cattle, and buffaloes. This disease can cause serious health problems and even death, resulting in significant economic losses for farmers. This study aims to develop an automatic detection system for Surra disease in horses using the K-Nearest Neighbor (KNN) classification method based on blood microscopic images. The research process includes image preprocessing stages (conversion to grayscale and resize), extraction of texture features using the Gray Level Co-occurrence Matrix (GLCM) method, and classification using the KNN algorithm. The extracted features include contrast, homogeneity, energy, and correlation. Model evaluation was carried out using a confusion matrix and evaluation metrics such as accuracy, precision, recall, and f1-score. Based on the results of the evaluation, the KNN model with a value of k = 1 was able to achieve an accuracy of 85%, with a high recall in the Infected class of 0.95, and an average f1-score of 0.85. These results show that GLCM and KNN methods are effectively used in detecting Surra's disease on blood images, especially in recognizing visual patterns that indicate infection. The developed system is expected to be a fast, efficient, and practical early detection solution in the field, especially in livestock areas such as East Sumba.