Lets assume that a robot is placed in an environment and some information about the environment (Map) is known. Also the robot does not know its exact position ( co-ordinates) in the given environment and it has to know its position. This process is call the Global Localalization. To solve this problem we now take the sensor(laser,sonar,etc..) readings of the robot in certain angles. Then we will go to each boundary cell of the obstacle and project the obtained sensor readings backwards along the corresponding angles. The point in the map with the highest number of intersections of these projections is the most probable position for the robot.