LUMPY SKIN DISEASE IDENTIFICATION IN CATTLE BY USING DEEP LEARNING APPROACH

Ethiopia is home to the largest livestock population in Africa, making the health and productivity of these animals a critical economic and food security concern. One of the significant challenges facing the Ethiopian cattle industry is Lumpy Skin Disease (LSD), a serious animal skin disease caused by the Lumpy Skin Disease Virus (LSDV). Early and accurate detection of LSD is essential for implementing effective disease management strategies. In this study, the researchers developed a deep learning-based system to automate the detection of LSD in cattle. They evaluated the performance of three prominent deep learning models: Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and VGG16, a pretrained CNN architecture. The results of the evaluation showed that the ANN model achieved an accuracy of 87.2% in correctly identifying LSD in cattle. The VGG16 model, leveraging its pre-trained feature extraction capabilities, demonstrated a strong performance with an accuracy of 97.2%. However, the standout performer was the custom-built Convolutional Neural Network (CNN) model, which achieved an impressive accuracy of 99%. The CNN's superior performance can be attributed to its ability to effectively learn and extract the distinctive visual features associated with LSD, enabling it to make highly accurate predictions. These findings suggest that the CNN-based approach is the most promising deep learning technique for automating the detection of LSD in cattle. By implementing this system, veterinarians and farmers in Ethiopia can enhance their ability to identify LSD early, leading to improved disease management, reduced economic losses, and better overall cattle health and productivity.