Details
| Original language | English |
|---|---|
| Title of host publication | Image and Graphics - 8th International Conference, ICIG 2015, Proceedings |
| Editors | Yu-Jin Zhang |
| Publisher | Springer Verlag |
| ISBN (print) | 9783319219684 |
| Publication status | Published - 2015 |
| Event | 8th International Conference on Image and Graphics, ICIG 2015 - Tianjin, China Duration: 13 Aug 2015 → 16 Aug 2015 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 9219 |
| ISSN (Print) | 0302-9743 |
| ISSN (electronic) | 1611-3349 |
Abstract
For the transmission of aerial surveillance videos taken from unmanned aerial vehicles, region-of-interest-based coding systems are of growing interest in order to cope with the limited channel capacities available. We present a fully automatic detection and coding system that is capable of transmitting HD-resolution aerial videos at bit rates below 1 Mbit/s. In order to achieve this goal, we extend the video coder HEVC by affine global motion compensation. Results of the computer vision algorithms control the extended HEVC encoder. For detection of moving objects, we analyze the video and compare a motion-compensated previous image with the current image. Image segmentation based on superpixels helps to select entire moving objects. In order to achieve low false-positive rates and low data rates, we use different motion-compensation algorithms for video analysis and video coding. Depending on the size of the moving objects on the ground, we can save up to 90 % of the data rate of regular HEVC without loss of image quality and the additional benefit of providing a mosaic of the video with moving objects.
ASJC Scopus subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Image and Graphics - 8th International Conference, ICIG 2015, Proceedings. ed. / Yu-Jin Zhang. Springer Verlag, 2015. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9219).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Region of Interest Coding for Monitoring the Ground with an Unmanned Aerial Vehicle
AU - Ostermann, Jörn
PY - 2015
Y1 - 2015
N2 - For the transmission of aerial surveillance videos taken from unmanned aerial vehicles, region-of-interest-based coding systems are of growing interest in order to cope with the limited channel capacities available. We present a fully automatic detection and coding system that is capable of transmitting HD-resolution aerial videos at bit rates below 1 Mbit/s. In order to achieve this goal, we extend the video coder HEVC by affine global motion compensation. Results of the computer vision algorithms control the extended HEVC encoder. For detection of moving objects, we analyze the video and compare a motion-compensated previous image with the current image. Image segmentation based on superpixels helps to select entire moving objects. In order to achieve low false-positive rates and low data rates, we use different motion-compensation algorithms for video analysis and video coding. Depending on the size of the moving objects on the ground, we can save up to 90 % of the data rate of regular HEVC without loss of image quality and the additional benefit of providing a mosaic of the video with moving objects.
AB - For the transmission of aerial surveillance videos taken from unmanned aerial vehicles, region-of-interest-based coding systems are of growing interest in order to cope with the limited channel capacities available. We present a fully automatic detection and coding system that is capable of transmitting HD-resolution aerial videos at bit rates below 1 Mbit/s. In order to achieve this goal, we extend the video coder HEVC by affine global motion compensation. Results of the computer vision algorithms control the extended HEVC encoder. For detection of moving objects, we analyze the video and compare a motion-compensated previous image with the current image. Image segmentation based on superpixels helps to select entire moving objects. In order to achieve low false-positive rates and low data rates, we use different motion-compensation algorithms for video analysis and video coding. Depending on the size of the moving objects on the ground, we can save up to 90 % of the data rate of regular HEVC without loss of image quality and the additional benefit of providing a mosaic of the video with moving objects.
UR - http://www.scopus.com/inward/record.url?scp=84943598035&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84943598035
SN - 9783319219684
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
BT - Image and Graphics - 8th International Conference, ICIG 2015, Proceedings
A2 - Zhang, Yu-Jin
PB - Springer Verlag
T2 - 8th International Conference on Image and Graphics, ICIG 2015
Y2 - 13 August 2015 through 16 August 2015
ER -