Global food security depends on maintaining the health of rice crops, and detecting diseases in rice leaves is essential to producing a strong harvest. Even though rice is a crop that is grown extensively, illnesses may provide serious financial difficulties. This work presents a unique approach that seamlessly integrates morphological and spectral information to improve the accuracy and robustness of rice leaf disease classification. Using the rice leaf disease dataset, our method uses Convolutional Neural Networks (CNNs) with transfer learning. Morphological features, encompassing observable and measurable physical characteristics of rice leaves such as shape, size, color, and texture, are extracted using deep learning techniques including edge detection, thresholding, and shape analysis. These measurements, including leaf length, width, aspect ratio, and surface texture, collectively contribute to a comprehensive morphological feature set, offering insights into structural variations associated with healthy and diseased leaves. Simultaneously, spectral features obtained through spectral analysis provide valuable information on the interaction of light with rice leaves. By measuring light reflectance or transmittance at various wavelengths, specific pigments or chemicals indicative of disease presence are identified. These spectral signatures facilitate differentiation between healthy and diseased leaves, synergizing with morphological information to significantly enhance disease classification accuracy. Our proposed method combines both morphological and spectral features to form a comprehensive classification model. Machine learning algorithms, including random forests, decision trees, support vector machines (SVMs), and neural networks, are trained on this joint feature set. Leveraging the strengths of both data types, our classifier accurately predicts the disease class of new, unseen rice leaves. A pre-trained convolutional neural network (CNN) model called VGG16 was fine-tuned using the rice leaf dataset in order to identify and categorize six main illnesses affecting rice leaves. With a remarkable 97.3% accuracy in illness categorization, our system outperforms previous machine learning algorithms, as shown by performance assessment criteria including accuracy, precision, recall, and F1-score. This integrated approach not only improves disease detection but also facilitates early intervention for effective crop management.