Drone video coding and communication

Energy efficient drone inspection system for spotting leakages

Unmanned Aerial Vehicles (UAVs or Drones) with conventional video and other sensors such as Infra-red (IR) cameras may be used for many applications. In collaboration with Zeuxion and Sky-Watch, the idea of this research project, supported by EUDP, is to use a drone based video and IR system to inspect energy systems. Further perspectives include to extend to multiview processing and systems.





The goal of the EUDP project is to develop and demonstrate a drone based system to inspect energy systems and distribution. A lightweight video and IR cameras system was developed, analyzed, and optimized as an on-board system. The drone based solution was demonstrated and tested and can be applied for example for district heating and transmission line energy loss inspection. The technical goal of the project is to make a lightweight automated drone capable of inspecting energy leaks in transmission lines, distribution networks, plants and field installations. This has a significant potential for saving energy and reducing CO2 emissions.

Using energy efficient drones for inspection and automating this process, three key goals are met: 1) Inspections becomes cheaper and more accurate 2) Energy leaks are more easily found and repaired 3) Even a minor reduction in overall energy loss represents huge savings.

Drone technology is at a stage where this is possible, and sensor technology is available in sizes and weights that allow for relatively lightweight drones to carry them. However this presents the challenge of communicating data from the drone to the ground and control commands back up.

The key technical outcomes, providing the basis for commercial success, are: 1) Designing a very efficient and safe video compression algorithm, 2) Implementing the compression algorithm in hardware suitable for aircraft and real-time use, 3) Designing the video and IR system incl. control system and transmission link to be power and space efficient,  4) Integrating this systems with sensors: IR and RGB cameras, and 5) Demonstrating and testing these for inspection of the energy systems.

For more information, you are welcome to contact Prof. Søren Forchhammer,, 4525 3622.



Related Publications

1. J. Søgaard, R. Zhang, S. Forchhammer, K. Hossain, Low complexity video encoding for UAV inspection, IEEE Proceedings of the 32nd Picture Coding, Nuremberg, 2016.

2. C. Mantel, and S. Forchhammer, Compression of Infrared images, IS&T Symposium on Electronic Imaging: Science and Technology. vol 2017, Visual Information Processing and Communication VIII, pp. 21-26, 2017.

3. E. Belyaev, C. Mantel, and S. Forchhammer, Low-complexity Compression of High Dynamic Range Infrared Images with JPEG compatibility, Proc. IEEE Visual Communications and Image Processing (VCIP), 2017.

4. E. Belyaev, C. Mantel, and S. Forchhammer, High bit depth infrared image compression via low bit depth codecs, Proc.  SPIE Optical Engineering + Applications, Infrared Remote Sensing and Instrumentation XXV, 2017.

5. E. Belyaev, S. Forchhammer, Drone HDR Infrared video coding via aerial map prediction, Proc. IEEE International Conference on Image Processing (ICIP), 2018.

6. K. Hossain, C. Mantel, S. Forchhammer, No Reference Prediction of Quality Metrics for H.264 Compressed Infrared Image Sequences for UAV Applications,  Proc. Electronic Imaging Symposium, San Francisco, 2018.



Søren Forchhammer
Professor, Group Leader
DTU Fotonik
+45 45 25 36 22