Abstract

The development of augmented reality games is challenging, especially of the ones of Superhuman Sports, where players are able to interact with smart augmented objects and to move freely. This thesis focuses on the detection and pose estimation of the object the players are interacting with, namely a flying drone, using OpenCV methods and the C++ programming language. We are dealing with challenging conditions, like varying lighting and varying distances between camera and drone. For a robust detection, the LED patterns extracted from the frames have to be corrected by eliminating the false positives and the false negatives and by introducing a mechanism which, in the case of a far away positioned drone, it will trigger the algorithm to look for the LEDs only on a small part of the image, where the drone is detected.

Results/Implementation/Project Description

Conclusion

[ PDF (optional) ] 

[ Slides Kickoff/Final (optional)]