Short of TV and film production, law enforcement has become the most video intensive of all professions. Contemporary law enforcement agencies must cope with terabytes and even petabytes of digital video data from numerous sources, including officer-worn cameras, in-car video systems, interrogation room cameras, jail cameras, civilian smartphones, the surveillance systems of businesses, the anti-crime cameras of local governments, and footage from broadcast media.
All of this video presents law enforcement with headaches. It's costly to store; it's costly to review, and it's costly to process for legally mandated release to the press and the public. And because the video captured of one single event can be in many different formats from numerous sources, it can require special technology to make it viewable.
Blurring the Picture
More than a few agencies have discovered after implementing officer-worn camera programs that the costs of the equipment and data storage are just the beginning of the budgetary hurdles they must overcome. They also need to be able to process the video for release to the court, the media, and the public.
Releasing video to the media and the public can be tricky and costly. This is especially true if the agency is required by law to do so. At least one state, Washington, currently requires agencies to make available all officer-shot video to the public by request. This has led to some serious complications for some law enforcement agencies in the state and a movement to change the law. But every agency in the United States is subject to Freedom of Information Act (FOIA) requests.
And a law enforcement agency can't just release video without taking steps to protect the innocent and at least some of its officers. Agencies have to be very careful of releasing footage of victims, witnesses, juveniles, and undercover officers. So the video has to be edited before it is released. Which can be labor-intensive and time-consuming considering that digital video systems capture images at a rate of 30 frames per second.
There was a time when editing law enforcement video required a skilled video editor running sophisticated software. But the makers of officer-worn video systems, including Digital Ally, TASER, Utility, and Vievu, have a better solution. They are now incorporating automatic redaction software in their evidence management software.
Using automatic redaction tools, the viewer can mark a target such as an object or person and the software will track that person or object throughout the video and blur it using a variety of techniques. For example, users can blur the entire person or object in what is known as a "Vaseline smear" or just part of the object. Such software can save an agency many hours of labor by either a sworn officer or a non-sworn editor. Only the copy of the video that will be released is redacted; the original remains unchanged to be used as evidence, if necessary.
Bob McKeeman, CEO of Utility, says redaction tools are a major concern for his company's customers and the automated redaction has to be 100% accurate. "You can't afford to miss a frame when the lives of witnesses and undercover officers are on the line," he says.
McKeeman says Utility tested its Smart Redaction software at last year's IACP conference in Chicago. A company employee wore Utility's Body Worn brand system and captured video of conference attendees waiting to enter an auditorium for a speech by President Obama, and then the video was redacted in a variety of ways using Smart Redaction. (You can see the video online at the Website www.smartredaction.com.)
Taking a Shortcut
One of the most time-consuming jobs in law enforcement can be reviewing surveillance video. When investigators don't know the exact time of an incident, they often have to watch hours and hours of this stuff to spot the perpetrators. But an amazing technology called "video synopsis," which can identify the pertinent information on a video and edit it to show just that data, may make long hours of tedious surveillance video watching a thing of the past.
Video synopsis was developed by a professor at the Hebrew University of Jerusalem and is now marketed by a company called BriefCam. Asked to explain how the technology works, BriefCam's General Manager for North America Amit Gavish wrote:
"Video Synopsis is the simultaneous presentation of events that occurred at different times. The technology separates dynamic moving objects (events) from static background and inputs the events into a database along with associated metadata such as color, size, speed, and other factors. When a given time-frame is called on the relevant events from that time are presented on-screen in non-chronological order. Additionally, should the user have some information about his or her target such as he wore red and went to the left, refinement tools can be applied so as to present only the events meeting those parameters. The original video can be accessed in a single click."
BriefCam Syndex software is in use by a number of law enforcement agencies in the United States and Canada.
Handling the Overload
The amount of video now being produced and processed by law enforcement agencies has been described as a tsunami. That's obviously an exaggeration when it comes to day-to-day operations. But an agency involved in a major incident can be quickly engulfed in video evidence.
After the Boston Marathon Bombing, the FBI and other involved agencies were inundated by news media, smartphone, and surveillance camera video. Fortunately, techniques for working with this diverse and abundant video had already been developed and proven effective by investigators and forensic video specialists working the Vancouver Hockey Riot.
After the Vacouver Canucks lost the 2011 Stanley Cup to the Boston Bruins in game seven played in Vancouver, some local hockey fans went on a rampage. The resulting riot injured 140 people and caused about $5 million in damages.
The public was outraged and the Vancouver Police Department and surrounding agencies formed the Integrated Riot Investigation Team (IRIT) to find out who participated in the riot and provide prosecutors with evidence.
Vancouver officials publicly requested that the public turn in any video of the riot. And people did, by the petabyte.
The IRIT investigators quickly realized that even though the Vancouver PD has perhaps the world's most sophisticated forensic video lab, this task was beyond their capabilities. The video was in dozens of different formats, dozens of different compression ratios, and used different encoding. So processing it for use by the IRIT investigators would require a monumental effort.
IRIT contacted the Law Enforcement and Emergency Services Video Association (LEVA), and it mobilized its Forensic Video Analysis Response Team. The video was sent to the National Digital Multimedia Evidence Process Lab at the University of Indianapolis. The Response Team included 43 forensic video analysts from numerous countries and 10 IRIT officers. They worked in three shifts, 24 hours per day, seven days a week from Sept. 26 to Oct. 9 to process the evidence using the lab's 20 workstations, an Avid ISIS shared storage solution, and Ocean Systems software such as dTective and Omnivore.
Grant Fredericks, technical manager for LEVA, says that Ocean Systems' Omnivore software was key to the success of the project. "That tool is magic," he says, explaining that it was used to convert all of the media into one format of uncompressed video so that it could be used by the IRIT investigators and the prosecutors. "We had hundreds of different sources and in about five or six steps we had to process it into something that the IRIT team could use," he adds. "It was quite a daunting process. There was even one analog video tape that we had to work with."
For More Information: