Vision based indoor navigation system using external camera mounted on room

Abstract

It is vital to have an excellent path planning for self-navigating mobile robots. Having considerations of the environment and objects in the space is vital in achieving a safe and effective path for robots. In indoor environments, objects and the environment are unpredictable, unlike that of in the road navigation where there are standard guidelines and road symbols. To overcome this issue, existing applications provide a solution to a certain extent. Indoor positioning systems based on Infrared, Wi-Fi and other similar techniques are existing in the industry. Also, there are solutions providing such systems with the use of a camera on the robot. Although such systems deliver the positioning of the robot and thus generating a path based on it, they fail to have an overall view of the entire space so that the generation of the path is more effective. To overcome this, the paper discusses here, a solution to the centralized vision-based navigation system for indoor robots. Our solution takes a view of the entire space beyond the immediate surroundings to the robot which was the only concern of previous solutions existing. By such a view, our solution can plan a path for the robot more effectively as it has all the objects in the space into consideration.