An extensive investigation of convolutional neural network designs for the diagnosis of lumpy skin disease in dairy cows.

Cow diseases are a major source of concern for people. Some diseases in animals that are discovered in their early stages can be treated while they are still treatable. If lumpy skin disease (LSD) is not properly treated, it can result in significant financial losses for the farm animal industry. Animals like cows that sign this disease have their skin seriously affected. A reduction in milk production, reduced fertility, growth retardation, miscarriage, and occasionally death are all detrimental effects of this disease in cows. Over the past three months, LSD has affected thousands of cattle in nearly fifty districts across Bangladesh, causing cattle farmers to worry about their livelihood. Although the virus is very contagious, after receiving the right care for a few months, the affected cattle can be cured. The goal of this study was to use various deep learning and machine learning models to determine whether or not cows had lumpy disease. To accomplish this work, a Convolution neural network (CNN) based novel architecture is proposed for detecting the illness. The lumpy disease-affected area has been identified using image preprocessing and segmentation techniques. After the extraction of numerous features, our proposed model has been evaluated to classify LSD. Four CNN models, DenseNet, MobileNetV2, Xception, and InceptionResNetV2 were used to classify the framework, and evaluation metrics were computed to determine how well the classifiers worked. MobileNetV2 has been able to achieve 96% classification accuracy and an AUC score of 98% by comparing results with recently published relevant works, which seems both good and promising.
Competing Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(© 2024 The Author(s).)