The increasing prevalence of crop diseases poses a significant threat to global food security, necessitating advanced solutions for timely and accurate detection. This research explores the efficacy of computer vision techniques in identifying and preventing crop diseases through the application of state-of-the-art Convolutional Neural Network (CNN) architectures: VGG16, ResNet50, and InceptionV3. Utilizing a comprehensive dataset of 50,000 images encompassing 15 crop diseases and healthy plants, the study evaluates the performance of these models based on accuracy, precision, recall, and F1-score. InceptionV3 emerged as the superior model, achieving an accuracy of 95.6%, precision of 94.7%, recall of 94.1%, and F1-score of 94.4%, significantly outperforming VGG16 and ResNet50 (p < 0.01). The integration of these models into a decision support system (DSS) and a mobile application facilitated real-time disease detection with an average response time of approximately 2 seconds. This rapid and precise identification system empowers farmers to implement targeted interventions, reducing dependency on broad-spectrum pesticides and promoting sustainable agricultural practices. The findings underscore the transformative potential of computer vision techniques in precision agriculture, enhancing disease management and mitigating crop losses. Future research should focus on improving model robustness through field testing and integrating additional data types to further refine predictive accuracy. This study represents a significant step towards leveraging advanced technologies for enhanced agricultural productivity and sustainability.