Strawberry plants, like many other crops, are vulnerable to a range of leaf diseases, which can cause substantial yield reductions. For crop loss to be avoided and high productivity levels to be maintained, timely diagnosis and precise classification of these diseases are essential. Conventional manual techniques for identifying diseases are frequently slow, error-prone, and limited in precision, particularly when recognizing early or subtle symptoms. To overcome these limitations, this study introduces an automated deep learning framework designed for precise classification of strawberry leaf diseases, employing advanced Convolutional and Recurrent Neural Networks (CNN and RNN) optimized with bacterial foraging and particle swarm optimization techniques. The framework combines CNN and RNN architectures, optimized to deliver high accuracy in distinguishing fungal and viral infections. By employing K-means clustering, Sparse Principal Component Analysis (PCA), and a CNN with ReLU activation, the model effectively extracts and refines features linked to disease. Further, region-growing segmentation and an optimized RNN enhance detection accuracy by capturing essential spatial and temporal patterns. Experimental tests using the PlantVillage dataset indicate a classification accuracy of 99%, highlighting the model's potential for early disease control and real-time agricultural surveillance. This strategy has a lot of potential to improve disease control methods, which will eventually increase crop productivity and yield.