Rice (Oryza sativa) production faces a severe menace from Blast disease, which appears as a result of Magnaporthe oryzae infection, thereby creating significant yield losses that threaten global food supplies. The successful treatment of these losses demands immediate leaf disease diagnosis followed by precise identification of affected areas. The proposed study develops an upgraded machine learning system that unites Convolutional Neural Networks (CNNs) with Support Vector Machines (SVMs) to enhance blast disease detection in paddy leaves. The preprocessed and resized dataset containing 72,000 images of healthy and diseased leaves was fed into a CNN model that obtained key features to be classified by SVM. The proposed model delivers better classification outcomes because it surpasses the results from traditional methods while reducing errors and improving sensitivity measurements. The main findings of this research consist of: The architecture uses CNN models together with SVM classification engines to reach improved performance results. The application of an enhanced image processing sequence for improving feature extraction effectiveness. Evaluation of model robustness across diverse environmental conditions. The experimental results show encouraging outcomes, which establish a base for potential agricultural field deployment. Additional research will investigate the adoption of this model with different crop diseases while also developing its performance for operational deployment.