Mobile Diagnosis of Foot and Mouth Disease in Cattle Based on Hybrid Deep Neural Networks

Foot-and-Mouth Disease (FMD) is a serious concern for livestock, especially cattle and pigs, as it endangers their health and can severely damage farming economies. Accurate and timely detection of FMD lesions in these animals is essential for disease management and control efforts. This study introduces a new method for detecting Foot-and-Mouth Disease (FMD) lesions by combining deep learning techniques. The proposed Hybrid Deep Neural Network analyzes image data to identify key features for accurate lesion detection.. The framework leverages the modified pre-trained CNN architecture of two models, i.e. MobileNet and NasNetMobile, to extract features from FMD lesion images because of their powerful image classification abilities. Through experiments conducted solely on image datasets, the performance of the proposed hybrid model is evaluated and compared with traditional CNN models. Out of the 800 images, the proposed hybrid model could correctly classify 781 images, which is 0.9763% accuracy, with a 0.9547% Recal, and F1 Score 0f 0.9768%. The results demonstrate the efficacy of the Hybrid Deep Neural Network in accurately detecting FMD lesions in cattle, highlighting its potential for improving disease surveillance and control in the livestock industry.