IEEE 2016 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET 2016)

237 / 2016-01-24 17:51:20

AUTOMOTIVE SYSTEM FOR VEGETABLE CLASSIFICATION

9225,5658,9226,9227,5604,9228

Efficient locating and detecting vegetables on the tree is one of the major requirements for the vegetables harvesting system .This work presents the vegetable detection using improved multiple feature based algorithm vegetable recognition is difficult because vegetables are deferrable objects that exhibit significant variations. The input images are the sections of the tree image. The algorithm is designed with the aim of calculating different weights of different features. These features represent the approximate location of the vegetable within an image. The proposed approach can be applied for targeting vegetables for robotic vegetable harvesting .The recognition is done by the SVM classifier based upon the statistical features derived from both color and texture features. The experimental results show, that the proposed system is significantly more accurate in identifying vegetables from the tree than the existing methods.