Analisis Perbandingan Klasifikasi Citra Genus Panthera dengan Pendekatan Deep learning Model MobileNet

Wildlife conservation is increasingly becoming a top priority as several species in the Panthera genus have experienced significant declines in their populations since the 1970s, due to illegal hunting activities, loss of natural habitat, and reduced prey. Their protection is therefore of paramount importance. In today's digital age, image processing and artificial intelligence (AI) technologies have changed the way we view and protect wildlife. In this context, the approach of using MobileNet models in deep learning, which is a branch of artificial intelligence, has proven to be very effective in overcoming complex challenges in image processing. However, despite MobileNet's potential in classifying images of the Panthera genus, not many studies have specifically compared it with other existing methods. Therefore, in this study, a comparative analysis of Panthera genus image classification using the deep learning approach of MobileNet model with alternative models from previous studies is conducted. The dataset used consists of 6,460 images with 6 labels: Jaguar, Leopard, Lion, Lioness, Tiger, and Snow Tiger, which are divided into training, validation, and testing sets. Based on the evaluation results, the proposed method using the MobileNetV1 model achieved the highest accuracy of 89.93%, followed by the MobileNetV2 model with 89.78%. This research is expected to provide valuable insights in the development of system implementation in an application on various platforms for image detection, to support species conservation efforts in the Panthera genus.