Enhancing Plant Pathology Detection and Stratification with a Tailored Mask R-CNN Framework

Our everyday lives are significantly impacted by the agriculture sector, which is crucial to the global economy. Despite its importance, a lack of knowledge in recognizing and controlling leaf diseases causes many farmers to have lower yields. It is crucial to quickly and accurately identify plant diseases since they have a direct impact on agricultural profit and loss margins. In order to identify and categorize plant leaf diseases, the Mask-RCNN model is used in this work to propose a novel solution to the problem. The first step in our technique is to eliminate noise from plant photos before feature extraction using the ResNet-50 model. The Regional Proposal Network (RPN) processes these properties after which they are divided and categorized using the Region-of-Interest (ROI) Align procedure. This study shows potential in changing disease detection in huge agricultural areas with a remarkable accuracy of 92.45%.By offering early warnings and interventions, these developments can be of considerable help to farmers and agronomists in assuring healthier crops and more effective agricultural methods. It is impossible to overstate the importance of identifying plant leaf diseases since it affects not only the agricultural industry but also related fields like biological research and agricultural institutes.