Plant species recognition using leaf contours

摘要

Plants have a significant impact on human's life and development. Plant species identification is essential for scientists working in the agricultural and environmental fields. Recognizing a species is not a trivial task due to the vast number of plant species that exist on earth. In this paper we present a computer-vision based technique to recognize plants using the contour of the plant leaves. The contour of a leaf first is extracted then processed and a feature vector is formulated. The feature vectors are stored in a database to be used for identifying plants in the worksite. For a scientist to recognize a plant, an image of the leaf is first feed to the computer, the feature vector is then extracted from the leaf and finally the species is identified by matching the leaf feature vector with the feature vectors stored in the database. The effectiveness of the proposed method is demonstrated by an experiment.