The study examines an automated method for identifying tomato leaf diseases using machine learning and image processing techniques. The proposed method uses Otsu's thresholding for image segmentation and extracts features using Local Binary Patterns (LBP) and Histogram of Oriented Gradients (HOG). These features are classified using a Support Vector Machine (SVM) with a polynomial kernel, achieving a classification accuracy of 94.6%. Bacterial, fungal, and viral pathogens cause extensive damage to tomato leaves, making an accurate and timely detection of these diseases essential for effective control and management. This methodology improves precision and lessens reliance on manual identification by combining effective preprocessing, segmentation, feature extraction, and classification techniques. The results highlight how sophisticated image processing and machine learning techniques may be used to solve problems in agriculture and enhance crop health management systems.