Deep Learning Based Lumpy Skin Disease (LSD) Detection

Abstract

The emergence of the lumpy skin disease has become a major threat to the livestock industry in recent years, causing high economic losses and health risks to both animals and humans. This virus is difficult to detect due to its complexity, making the early detection and accurate diagnosis of this virus essential. This study will explore the utilization of convolutional neural networks (CNNs) to efficiently and accurately detect and identify the LSDV than traditional methods. Further, the advantages of using CNNs for this purpose has been discussed and some of the applications of this new technology has also been explored. Additionally, the future potential of using CNNs to perform virus detection is also discussed. However, Lumpy disease is classified differently based on its severity. To determine the extent to which the animal is impacted by lumpy skin disease, it is necessary to recognize various stages of the disease. This research study referred to the use of several CNN architectures and Regression algorithms to detect the Lumpy skin disease virus as early as possible. The architectures explored are and the EfficientNet-EfficientNetB7 architecture, MobileNetV2, EfficientNet-EfficientNetB3 architecture, VGG16, InceptionV3, ResNet50, VGG19, Xception and DenseNet201. The paper thoroughly describes all of the steps required to carry out the disease detection model, from data collection to process and outcome.