Abstract:Classification of medicinal plants is a challenging process through the automated system and achieving proper result is a rigorous work. India is well known for its prosperity of medicinal plants and its medicinal practice. In this modern world a person may know few plants which are common in place. There are wide varieties of plants which are unaware. To come out of this problem and to make use of all the medicinal plants an automated system is useful. The system can be used by researchers, students and in the medicinal production sector. The plant has its own properties and uses, preserving such plants makes very helpful for the future. The existing system also helps in classifying the plants and our proposed study helps in classifying the plants with lower quality images whereas the existing asks for the high-resolution images. One of the major goals of the study is to create native dataset using low cost capturing efforts. Proposed work contains contour-based segmentation which deeply considers leaf morphology, feature extraction using Local Binary Pattern and Wavelet methods and classification using supervised K-NN classifier.

Medicinal Leaves Recognition Using Contour-Based Segmentation