Classification of Medicinal Leaf by Using Canny Edge Detection and SVM Classifier

Plants play an indispensable role in human life by providing food, oxygen, medicine, shelter, fuel along with environmental protection. Besides, lot of plants contain medicinal qualities and active ingredients for treatment. In modern years, many useful plant species have become vanished or have been destroyed due to factors such as population growth, global warming and lack of government support for research activities. Scientific research in the field of machine learning and image processing has concentrated on classification of medicinal plant species. Plants are recognized and classified by their leaf features. We assort leaves based on their shape and size. An automated system is presented in this paper which enumerates various properties of leaf shapes such as width, length, area of leaves, area of rectangle, perimeter of leaves in pixels. In the proposed work, we captured an image of the leaves and convert colored images into grayscale and binary image and further resized in order to process it rapidly and calculate the prominent features. Two machine classifiers K-nearest neighbor and Support Vector Machine are analyzed to regulate the optimum accuracy for classifying the plants.