A Leaf Recognition Technique for Plant Classification Using RBPNN and Zernike Moments

Abstract: Plants are among the earth's most useful and beautiful products of nature. Plants have been crucial to mankind's survival. The urgent need is that many plants are at the risk of extinction. About 50% of ayurvedic medicines are prepared using plant leaves and many of these plant species belong to the endanger group. So it is indispensable to set up a database for plant protection. We believe that the first step is to teach a computer how to classify plants. Leaf /plant identification has been a challenge for many researchers. Several researchers have proposed various techniques. In this paper we have proposed a novel framework for recognizing and identifying plants using shape, vein, color, texture features which are combined with Zernike movements. Radial basis probabilistic neural network (RBPNN) has been used as a classifier. To train RBPNN we use a dual stage training algorithm which significantly enhances the performance of the classifier. Simulation results on the Flavia leaf dataset indicates that the proposed method for leaf recognition yields an accuracy rate of 93.82%

Keywords: Zernike moments (ZM), Dual stage training algorithm, Radial basis probabilistic neural network, Gray Level Co-occurance Matrix.