HYBRID LION OPTIMIZATION WITH FASTER MRCNN TO CLASSIFY PLANT LEAF DISEASE

Abstract

Plant leaf disease creates disaster on agricultural production and crop output. Reliable and prompt disease categorization is critical for efficient disease control and protection of crops. In this research, proposed a hybrid approach that combines Hybrid Lion Optimization (LO) with Faster Mask R-CNN to classify plant leaf diseases. Lion Optimization leverages the hunting and social behaviours of lions in nature, combined with genetic algorithms, to enhance the limitations of the Faster Mask R-CNN model. By applying LO, attempt to discover the best collection of hyper parameters or architecture configurations that make the most of the performance of the leaf disease classification system. The Faster Mask R-CNN, an advanced object detection and instance segmentation algorithm, serves as the core framework for disease classification. It extracts data using a deep CNN landscapes from plant leaf images and performs region-based detection to identify and classify diseases within those regions. The consequences of our experiments determine that the hybrid approach improves the accuracy and efficiency of classification. The hybrid optimization algorithm effectively classify plant leaf disease with accuracy 97.4%. This hybrid approach has significant implications for the field of agriculture, offering a reliable and automated solution to aid in the diagnosis and categorization of plant leaf diseases.