Low-cost super-resolution algorithms implementation over a HW/SW video compression platform
EURASIP Journal on Applied Signal Processing
| Hi-index | 0.00 |
As video surveillance systems migrate from analog to digital, law enforcement agencies are entering a new era of forensic image analysis in which storage and enhancement can be performed exclusively on a computer. With a fixed compression ratio of 5:1, native digital video (DV) is far from perfect. However, it is inevitable that networks of DV surveillance cameras will soon be utilized in corporations, government facilities, and airports throughout the world. The keys to expediting this conversion include convincing the law enforcement community that the DV format is indeed acceptable as a replacement for analog NTSC video captured within closed circuit television systems, and that, in the future, enhanced forensic digital video will be admissible as evidence in most courts of law. Empirical results are presented, in which super-resolution video enhancement is applied to similar DV and digitized NTSC fields in order to compare the high-resolution video stills that result from multi-field integration and upsampling. Visually, it is shown that native DV has excellent quality compared to digitized NTSC video, and that super-resolution enhancement can indeed be useful for extracting additional information from either format.