CatEarMites: An Approach of Detecting Ear Mites of Cat Using Convolutional Neural Network

Machine learning and deep learning are two research fields that have recently gained popularity. Predictions made using these approaches are highly appealing these days. Pet illness prediction is a relatively new concept. In this field, there is virtually little study. Ear Mites is an illness that many pets suffer from, specially, cats. We have developed a program that can predict this condition based on photographs of a cat's ear. We've divided the images into two categories. The first class is ‘infected’, the other class is ‘not infected’. The study's major purpose is to determine whether or not a cat is afflicted with ear mites. The dataset has been accurately gathered, classified, and processed to prevent conflicts. A standard deep learning classifier, the Convolutional Neural Network (CNN), was utilized to get a greater accuracy rate. This model has an accuracy of 88 percent. One of the most sophisticated image identification technique is convolutional neural networks. The findings reveal that CNN-retrieved high-level features which are more discriminative and effective than typical hand-crafted features like LBPH and Haar-WT (Wavelet Transform). As a result, our CNN model is a top-performing technique for identifying ear mite disease and might be employed in real-world situations.