Method of identification of foliage from plants based on extraction of multiple features of leaf images.

The extraction of features of images of plant leaves is important for automatic classification and identification of plants. In this study, plant leaves were used as research objects. Through synthetic extraction of color, shape and texture features of leaf images, a method for automatical classification and identification of plants is proposed based on image analysis of a SVM (support vector machine) principle. After training and recognizing images of fifty parts of foliage of plants, our classifier has achieved good performance with a recognition rate of 91.41%,compared with the results of the BP (Back Propagation) neural network and the KNN (K-Nearest Neighbor) identification method.