Carrom Playing Robot with Automatic Shot Selection

Abstract

Robots are most innovative and influencing these days, nearly emerging in every field of science. To understand its reach to automation we are aiming to make a robot which can play carrom like human by using artificial intelligence. Most important feature of the robot is the visual recognition system which help the robot to take input from surrounding. The visual recognition system is the primary and most important feature as it takes the input from surrounding and gives it to the brain that is Microcontroller board with the help of camera. It works on computer vision library in python OpenCV, by using this library we build an algorithm to identify the coin on the carrom board. The algorithm made can also identify the coin on the board is black, white or Red (colour identification) and whether it is a coin or a striker. We have built a mechanism by using geared motors, servo motor, shafts and joints which strike the coin on the board precisely. Here while making the robot we tried to mimic the human wrist movements but with constrained motion. To handle all the information and programming we used an ad a fruit metro microcontroller board which is the brain of the entire robot.