On a global scale, potatoes are among the most extensively cultivated crops and serve as a fundamental food source in numerous areas. As one of the most versatile vegetables, potatoes are essential to cooking. Nevertheless, certain diseases that have the potential to significantly impede agricultural productivity and growth can impact potato crops. The objective is to create a technologically advanced and effective disease detection system that will improve potato production and digitise the diagnosis process. The major objective of this work is to detect potato illnesses via the analysis of leaf photos. The implementation of the MobileNetV2 model will simplify this challenge. The present work introduces an automated deep learning system that employs image processing methodologies for the purpose of detecting and classifying leaf illnesses. Experimental evidence has shown that image processing is a valuable approach for identifying and understanding these disorders. The dataset collected almost a thousand images of potato leaves affected by early blight (EB) and late blight (LB), as well as healthy leaves, from Kaggle for this study. This concept employed several pre-trained models to detect and classify these images. Specifically, the MobileNetV2 model achieved a prediction accuracy of 95.22%. In the context of potato disease identification, our results indicate that MobileNetV2 outperforms other existing methodologies.