Red-light traffic cameras present one of the most controversial uses of public surveillance video. Charles Hintz has researched the topic and testified for the defense in law courts on numerous occasions. Charles points out he is not a lawyer and is not offering legal advice, although he has appeared 11 times as an expert witness and been on the winning side in 8 of those cases in Bay Area courts. As an expert technical witness in red-light litigation, he has numerous insights into this type of video evidence. He states: “There are two kinds of automated video-enforcement systems: those that have lost class-action lawsuits and those that, in time, will lose class action lawsuits.”
What kind of image processing is used in red-light automated enforcement?
Charles will describe the uses of MPEG-1 for motion images and JPEG for still images of alleged violators. Red-light systems use lossy MPEG-1 compression because the transmission pipeline back to headquarters has such limited data capacity. Question: to what extent is the video evidence damaged by the extreme compression of data? That's the key to understanding the controversy surrounding this use of video technology in law enforcement. [see www.videohelp.com/vcd ]
Although the CCD camera at the intersection may shoot at a standard resolution of 768x480 pixels, only 352x240 (CIF: Common Interface Format) or one-fourth of full resolution (ITU-B 601) or 25 megabits per second (Mbps) are processed. A T-1 line only carries 1.5366 Mbps from the intersection back to the red-light-camera company for further processing -- technical and administrative -- before submission of the evidence to local authorities for prosecution. The compression and transmission process of the original footage with a total running time of usually 12 to 20 seconds reduces the motion image data rate to less than 1 Mbps. The system uses the remaining 408 Kbps of T-1 capacity to carry high-resolution still images that purport to show clearly the car's license plate and the face of its driver. The motion video purports to depict faithfully and accurately the traffic incident. [see www.crimevision.net/Manuals/LEVA_BestPractices.pdf ]
In addition to describing the technology behind the systems, Charles will discuss how, under California law, some courts have found red-light imaging evidence inadmissible owing to “spoliation” of evidence -- literally the significant damaging of evidence -- that has caused a number of jurisdictions to lose red-light-camera cases. [see californiadiscovery.findlaw.com/spoliation.pdf ]
Charles says red-light camera technology has a number of weak points. The SMPTE Time Code (source time code) and the time/day stamp on the video is sometimes not maintained correctly (chain-of-custody), with seemingly no reference to SMPTE Absolute Time Code (ATC) or the SMPTE Epoch (time zero) based on the time -- 00:00 -- on 1 January 1958 in White Plains, New York, that the code was officially standardized. [see www.smpte.org/standards/tf_home/S404M_DRAFT.pdf ]
“Stream-based codecs are sometimes considered inappropriate for forensic purposes because they have the characteristic of combining information from different frames and different points in time to build an interim frame that represents only what should be happening at that point, not what's necessarily happening.” [see www.broadware.com/files/ACodecMoment.pdf ]
Charles states, “MPEG compression is another layer of deception, created and refined to bring the theatrical experience into the home. Although we have the impression we're viewing the high-resolution film, 97.5% of the original film information has been discarded during compression. The injury to the quality of the film resolution is masked by an intentional layer of deception.”
Through various forms of video surveillance, has “Big Brother” arrived? This SMPTE meeting may offer a look into the future. The San Francisco Section plans to revisit periodically the issue of video in surveillance. “Advanced imaging” -- or sometimes, not so advanced -- no longer remains the exclusive province of broadcasters, program distributors, and videographers!
Welcome: SMPTEsf welcomes members and friends to attend without charge. If you don't have a ticket, you will not be admitted to this meeting. No substitutions.
Good news, though: if you can't attend the meeting, that's no problem -- we're now webcasting most of our meetings, including this one. Just a few days after the event, we'll all be able to watch the edited show online as a streaming video or an on-demand download on www.smpte.org and hopefully via YouTube.com and/or other outlets. We'll send everyone details soon by email.
See you either in San Leandro or online!
Peter Hammar
Chairman, SMPTE San Francisco Section