Infrared Camera Based Indoor Positioning System

Ng, Jun Ming (2017) Infrared Camera Based Indoor Positioning System. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Ng Jun Ming Full Text.pdf Restricted to Registered users only Download (2MB)

Abstract

The project design titled as “Camera Based Indoor Positioning System” focuses the key issues of indoor coordination for automotive robot. Indoor Positioning System is known as IPS which is a system to track the location of users or objects within indoor area. With the increase in mobile computing devices and image processing technique, it has become one of the important method for location tracking system at any point in term of safety and providing accurate results. Therefore, the scope of this thesis is to research and implement a new method for indoor positioning system that provide lower cost and higher accuracy. There are many similarities in existing IPS to that of creating new IPS with using different type of sensors. By analysing different technologies and algorithms of other IPS, also considering the trade-off between accuracy and cost, this thesis has concluded that , camera based Indoor Positioning system could be one of the most promising approach in determine the indoor location of users/device in the future. The scope of this thesis is to determine the location of device or users inside the building with better accuracy using camera with image processing techniques. In this thesis, the IPS will be using new method to detect the infrared signal which is using normal webcam as receiver instead of common infrared sensor and the infrared signal will be transmit from the customise infrared LED beacons. Location of the users can be determined by identify the changes of pixels in the image capture from the camera. Therefore, this project is required lots of image processing techniques such as contour detection, nearest neighbour algorithm, multiple objects detection and etc. Several experiments were run in offline phase to determine the accuracy and performance of the Camera based Indoor Positioning System. Experimental results show that Camera based IPS is highly efficient in both accuracy and costs. For instant, the offset between the actual position and estimate position is only up to +/-1.4 cm for a testing of total travel of 1.5m. As for future works, the accuracy of location could be increased as the resolution of camera is getting higher because the higher resolution which led to better sensitivity of the IPS (1 pixel represent smaller distance).