Image Classification System through the application of Deep Learning for the pre-diagnosis of bovine dermatological diseases

Abstract

The project addresses a comprehensive scenario within the livestock industry, focusing on the development of an Image Classification System using Deep Learning for the pre-diagnosis of Bovine Warble Fly, Contagious Nodular Dermatitis, Dermatophytosis, and Dermatophilosis. These dermatological diseases represent significant challenges to cattle health and, by extension, to public health and the economy of the livestock industry. The proposal aims to apply Deep Learning, specifically by combining Convolutional Neural Networks and Multilayer Perceptrons, to analyze images and speed up the diagnostic process. This approach not only facilitates early treatment, preventing the spread of diseases between animals and humans, but also offers an efficient solution to expedite the pre-diagnosis of various bovine dermatological conditions. Preliminary results indicate that the project could become a valuable tool to mitigate the impacts on animal health and the livestock sector as a whole.