Reality Check

Oct 1, 2005 12:00 PM, BY SAM BURGISS

Video is all around us. The amount of video used as evidence in investigations is increasing. The London bombings yielded video that was analyzed, enhanced and led to the information about the perpetrators of the violence. All this evidence is in one of two forms — analog, which is known as VHS tapes, and digital, a relatively new way to store video surveillance. There is a debate in the field of forensic video processing that centers around the new digital formats used for capturing and storing surveillance video. This digital debate involves issues such as admissibility of video evidence into court. With an understanding of the fundamentals of digital video, law enforcement and security agencies can make better, more informed decisions.

Advances from analog to digital in various consumer devices suggest that “digital” means “better” or “higher quality.” But digital is no panacea. In many cases digital is better and equates to higher quality. However, digital video's usefulness in forensic examinations is limited because the technology requires file compression, which can compromise quality. Video forensics is still evolving, and since standards do not exist, true digital video proof can be questioned.

Analog is the industry stand-by because video recordings uses a standardized format developed by the National Television Standards Committee, referred to as NTSC format video. Because there is a standard format, its quality is better understood and accepted in the realm of video forensics. No such standard exists for the digital video formats of CCTV and video surveillance systems, thus acceptance is significantly lower. DVR manufacturers use various proprietary formats, thus making video analysis and post-processing far more challenging. The varying characteristics of these digital formats leads to questions concerning the quality and consistency of these videos. The industry continues to argue about whether digital video reliably and accurately represents the scene captured by a camera.

Without compression, digital video is too large for everyday use, especially when considering its use with daily surveillance footage. Uncompressed digitized NTSC video, with 720 pixels × 480 lines, eight bits/pixel per color, three colors and 30 frames per second (a common description of digital video), requires 248 Mbps, making one minute of uncompressed video 1.8 GB.

Let's put this into perspective using some of today's standard technology. For instance, a CD-ROM can hold only 20 seconds of uncompressed video. A DVD can hold only two minutes. If DVDs used uncompressed video it would be necessary to change disks more than 40 times to watch a 90-minute movie. One would need to download 3.72 GB of uncompressed video to get one two-minute movie trailer. Using dialup it would take 6.6 days; on a cable modem, it would take 1.5 hours.

Theoretically it is possible to compress video without losing any signal quality. This compression is referred to as lossless. In practice, this type of compression is never used because it does not reduce the amount of data enough to make video usable.

The original motivation for compression was data reduction. Consumers wanted digital video, but it could not be captured in its original form. The inventors of digital compression simply considered how to make the video “look the same” while reducing the amount of data. When a person watches TV, a movie or a video on their computer, they simply want a reasonably sharp image, — and compression can achieve that.

However, video forensics processing has a different goal — focusing on the details. Zooming in on a t-shirt and seeing the detail of the t-shirt logo is not important to consumers, but can break the case in forensics.

To reduce the amount of digital data, video compression techniques reduce the amount of detail in the images that make up the video. Other techniques reduce the amount of repeated or “static” information in the video. The list of techniques goes on, but they all serve to reduce data, which serves to compromise the post-processing outcome. Some data somewhere is lost. In essence, parts of a frame are “cut out” to reduce the space needed to store the image.

In most cases, the portion that is removed is intact on another frame. For example, in the video of a room, some elements from frame to frame are unchanging and, therefore, can be retained in just one frame and removed in subsequent frames. But the selection of what pieces are removed is subject to the algorithm. If the compression algorithm is “messy,” it can “punch holes” in the important things — herein lies the center of the debate. Can video evidence be fully trusted when small portions of it are missing? Until an industry standard is approved, agencies have to make their own choices.

With all these issues to combat and so many questions raised about digital video, agencies must wonder what value video evidence brings. Just as DNA evidence met its share of controversy in the early 1980s, likewise, digital video is here to stay and bound to be a permanent fixture in the world of surveillance and security. The industry is charged with establishing a recognizable standard for capturing and compressing digital video, which will in turn support the consistent processing of video for forensics and encourage further acceptance in the court system. Until standards are established and digital video technology becomes more widely accepted, agencies are challenged to devise their own approach to digital including agency-wide accepted “best” practices for dealing with this technology.


Sam Burgiss is director of products at Signalscape Inc., Cary, N.C., (signalscape.com), a provider of advanced signal processing solutions. The company's digital signal processing solutions are incorporated into hardware and software technologies for the law enforcement and security agency markets.

Want to use this article? Click here for options!
© 2010 Penton Media Inc.

Story Missing Your Link?

Is the above story missing a link? Is it missing a link to your company, or your website? If this is the case please e-mail us and we'll add the link as soon as possible. Thank you!

Latest Issue Cover

This Month in Govt. Security

Popular Articles

Resource Center

Webinar

A Cost-Effective Framework For Total Security Integration

Join AC&SS and MAXxess as they review two different IP-framework applications
Wednesday, July 30, 2008 at 2:00pm ET/11:00am PT

Register Now!

Latest News from Access Control

Back to Top