Computer Vision on UAVs – practical considerations (Talk)
- Eric Price
Computer vision on flying robots - or UAVs - brings its own challenges, especially if conducted in real time. On-board processing is limited by tight weight and size constraints for the electronics while off-board processing is challenged by signal delays and connection quality, especially considering the data rates required for high fps high resolution video. Unlike ground based vehicles, precision odometry is unavailable. Positional information is provided by GPS, which can have signal losses and limited precision, especially near terrain. Exact orientation can be even more problematic due to magnetic interference and vibration affecting sensors. In this talk I'd like to present and discuss some examples of practical problems encountered when trying to get robotics airborne – as well as possible solutions.
Biography: Eric Price started programming at age 11 and continued on this path that led to his graduation at University of Stuttgart, with a major in computer vision and machine learning. During his studies he got involved in a number of international open source projects - most notably OpenPilot, where he took responsibility for the simulation framework, sensor fusion and flight control algorithms. His thesis fully plunged in the field of unmanned aerial vehicles, investigating algorithms for on board SLAM (simultaneous locating and mapping). Since graduation in 2013, Eric Price has been working as a system developer for TGU Smartmote, a spin-off from the materials testing institute in Stuttgart, where he develops the soft- and firmware for cloud backed distributed wireless sensor networks and participates in various research projects.
