Optimized Deep Learning Framework for Early Detection of Rice Leaf Diseases Using CNNSVM and FRRSALeNet Models

The early detection of rice leaf diseases is crucial for enhancing crop management and ensuring food security. This study presents an optimized deep learning framework for the early detection of rice leaf diseases using a comprehensive dataset that includes images of various disease types, such as Bacterial Leaf Blight, Brown Spot, Healthy, Leaf Blast, Leaf Scald, Narrow Brown Spot, Neck Blast, Rice Hispa, Sheath Blight, and Tungro. Two novel hybrid models were developed and evaluated for this task: a CNN-SVM hybrid model and a Fractional Remora Reptal Search Algorithm (FRRSA)-LeNet hybrid model. The CNN-SVM hybrid model combines the feature extraction capabilities of Convolutional Neural Networks (CNNs) with the classification strengths of Support Vector Machines (SVMs), resulting in enhanced accuracy and robustness in disease detection. On the other hand, the FRRSA-LeNet hybrid model leverages the efficiency of LeNet for image classification, coupled with the Fractional Remora Reptal Search Algorithm for optimized training, achieving improved convergence and accuracy. Both models were trained separately and evaluated based on classification performance, demonstrating their effectiveness in detecting rice leaf diseases with high precision. The proposed framework offers a promising solution for automated disease detection in rice fields, potentially aiding farmers in early intervention and effective crop management.

Keywords: Rice leaf diseases, CNN-SVM hybrid model, LeNet, Fractional Remora Reptal Search Algorithm, deep learning, disease detection, image classification, precision agriculture.