Thai Fruit Recognition System (TFRS)

Abstract

The main objective of this research is to develop a computer system that can recognize Thai fruit by using image processing techniques. Our system is called "Thai Fruit Recognition System (TFRS)". The system consists of four major components, which are: 1) image acquisition, 2) image preprocessing, 3) recognition, and displaying result. We used three important features to recognize Thai fruit, which are RGB colors, size, and edge shape. The TFRS is trained system with 39 kinds of Thai fruit on 312 images. We test our system on 123 images of 39 kinds of fruit for a training dataset and 30 images of 10 kinds of fruit for an un-training dataset. The precision of our system is 93.50 percent for the training dataset and 86.67 percent for the un-training dataset. The average access time is 54.68 seconds per image. Copyright 2009 ACM.