Buddhist Amulet Recognition System (BARS)

Abstract

There are many kinds of Buddha amulets, which are difficult for people who want to recognize all of them. The objective of this research is to build the computer system that can recognize some Thai Buddhist amulets. Our system is called -Buddhist Amulet Recognition System or BARS¿. It consists of 4 main components; 1) image acquisition, 2) image preprocessing, 3) image recognition, and 4) display result. In image acquisition component, the Buddhist amulet picture is taken by a digital camera with a white paper for the picture background. In image preprocessing component, the Buddhist amulet image is enhanced for improving a recognition precision rate in the next stage. In recognition component, a template matching technique is applied to recognize the Buddhist amulet image. In display result component, a graphic user interface (GUI) is created for displaying the recognition results. We have collected 52 kinds of Buddhist amulets with the total of 318 images for the training data set. We determine correlation value equal to 500 and this program precision equal to 80 percent and take 0.76 milliseconds per image.