<p>Plant health in agriculture is very important for crop productivity and food security. Conventional methods of plant disease detection are manual, requiring loads of time and labor, and are often subjected to human error. Considering that the visual manifestations of leaf disease are one of the basic symptoms of diseases, which can be promoted or detected first on leaves, the focus of major research has shifted toward leaf disease detection using computer vision. In computer vision, deep learning models achieved a superior performance showing an ideal solution to plant diseases diagnosis from an image-based leaf analysis. A detailed study was conducted from 2015 to 2025, which highlights the features of advanced techniques in plant leaf disease detection. Under this study, a systematic review and evaluation of advanced machine learning and deep learning methods which are used for recognition and classification of plant leaf diseases is conducted. Amid all the models, convolution neural network is the most accurate model in detecting plant leaf diseases. The major&#xa0;focus of this review is to guide the researchers to identify the most efficient approaches&#xa0;for plant leaf disease detection. It also lists the advantages and drawbacks of current models that are based on these models. Nine major research inquiries were prepared to govern the review’s scope; these questions covered aspects including datasets, image preprocessing, image segmentation, feature extraction, and classification techniques. It concludes by providing relevant answers to the research inquiries posed, in addition to future recommendations for plant leaf disease detection and classification.</p>