Determining distance and shape of an object by 2D image edge detection and distance measuring sensor

In modern robotics path finding is usually done by storing 3D models of objects into robots memory so the robot can compare the object in front of the camera with the model in its memory while it is trying to navigate between obstacles. With this approach we need 3D model of every object the robot may come into contact with, which is possible only if we confine the robot to a known environment. This paper proposes that path finding for mobile robots be based on 3D extraction of a 2D image edge detection and infrared (IR) distance measuring sensor.