Our project focuses on indoor positioning, where no GPS signals can be used.
Due to the width of this subject, we limited ourselves to deal only with a specific method of positioning, namely – image processing, and to try to enhance performance only in terms of accuracy and speed (and not, for example, coverage as much space as possible, accessibility and ability to be implemented in various of places or any other aspect).
Our goals in the project were learning and experiencing new workspaces, mainly which are related to image processing, and developing a system which can recognize and position a few objects simultaneously, while moving.
The system basically is an application which includes:
A camera which sends pictures to be processed, the camera pictures the rooms continuously and sends frames one by one, we used a smartphone camera.
A QR code, stamped on the objects – used for understanding were the objects are, and is written according to the features that we want to support in (for example – different groups). We wrote the QR code in python, using dedicated libraries for QR.
A program which extracts the information of the objects and calculates their exact position and angle, after calculating the results are transferred to the DB.
Robots that are moving around inside the room according to predefined conditions (when and where to go, how fast and when to stop). We used robots of EV3 and wrote the equations of the movement in LabVIEW.
And a screen which is used for presenting the movement of the different robots. The results of the calculation explained above – means, the id, position and angle – are read from the DB.
The system we developed is able to detect a number of objects in the same time – the specific number is depended on the height of the camera and the lighting conditions, and follow their movement accurately.