End-to-End Semantic Leaf Segmentation Framework for Plants Disease Classification

Pernicious insects and plant diseases threaten the food science agriculture sector. Therefore, diagnosis detection of such are essential. Plant disease classification is a much-developed research area due to enormous development in machine learning (ML). Over last ten years, computer vision researchers proposed different algorithms for identification using ML. This paper proposes an end-to-end semantic leaf segmentation model identification. Our uses deep convolutional neural network based on (SS). The algorithm highlights diseased healthy parts allows affecting specific leaf. successfully foreground (leaf) background (nonleaf) regions through SS, identifying as parts. As label provided by method each pixel, information about how much affected also estimated. We use tomato leaves test case our work. CNN-based publicly available database, PlantVillage. Along with PlantVillage, we collected dataset twenty thousand images tested framework it. obtained average accuracy 97.6%, which shows substantial improvement performance same compared previous results.