The world is a place for many living things apart from human beings, like plants, trees, animals, bacteria, insects, mammals, reptiles, etc., in which plants and trees are the vital source for oxygen for us humans. Every organism has its own life-time where it could live prolong for several hundred/ thousand/ billion years like a 'Bristlecone pine-tree' or algae like 'Proterocladusantiquus'. During their lifespan each plants and trees also experience sickness or illness due to diseases which results in withering away, fruits and leaves drop. Studying these factors affecting the plants, plant's lifespan, solutions and classifications of plants and diseases are known as “Plant Pathology”. Though the Plant Pathology could be achieved through machine languages and automated machine approaches, the involvement of humans for classifying and categorizing diseases have been the only approach till-date. It is costlier, time consuming and labour intensive. Hence the proposed research aims at developing an algorithm that could automatically identify, classify and categorize the plant inputs through RBFNN (radial basis function neural network) with image segmentation through weighing function, where the optimization is done through PSO aiming at efficient and higher accuracy rate based rapid results. The developed RGM (region growing algorithm) increases network efficiency for speed and clustering the common attribute-based seeds towards extraction process of plant's feature. By focusing on fungal diseases classification with factors like, leaf spots, leaf curl, late blight, common rust, early blight and cedar-apple rust the study was carried-out. The developed algorithm and outcome through test and train method shows efficiency, accuracy in classifying and categorizing the plant diseases.