SF surveillance cameras don’t stop violent crime
A study conducted by the UC Center for Information Technology Research in the Interest of Society reveals that the surveillance cameras installed around San Francisco have failed to deter violent crime. They have been effective, however, in lowering property crimes within 100 feet of the cameras by 24 percent. But before you conclude that all surveillance video projects are useless based on the SF study, read the report. It points out several problems:
- disorganized city management of the surveillance project: strangely enough, it’s not the SF Police Department in charge of this, rather, a melange of agencies are overseeing the project;
- attorneys and police officers not trained to use the footage;
- poor quality of the video: SF video cameras are set to film only three to four frames per second (movies shoot between 25 to 30 frames per second);
- people committing crimes are using disguises: what a surprise, dress up in a hoodie, put on a hat, wear a mask, anything to cover your face.
Download the San Francisco video surveillance study (if this link does not work, please use this).
Related stories:
Last year I posted a story questioning the effectiveness of wireless video surveillance and in particular, mentioning a UK report that says CCTV cameras in the UK have not been effective against crime:
Wireless video surveillance gains popularity, but it is effective?
- – - – - – - -
*Advertise now on Muniwireless.com. Download our 2009 Media Kit and contact us.
*Join the MuniWireless / Muni WiMAX Linked In Group for networking and discussions. Over 200 members as of 19 December 2008.
*Sign up for the MuniWireless / WiMAX newsletter.
Related posts:
Loading ...
I really want to make it clear from the post above that many of the disappointments with the SF project are distinct to San Francisco. Please download and read the report. It’s long (184 pages) but truly worthwhile if you are planning on installing wireless video surveillance, or if you already have a network running and need to find out what you can do to improve the results.
Despite its inefficiencies, I don’t think the San Francisco municipal surveillance system is a disaster by any means. Given ample funding, the city actually has the opportunity to operate one of the only successful municipal surveillance systems out there. Granted, a few major functionalities must be put in place for this to happen, but it is a project that is worth the time and effort.
I wrote a bit more about adjustments necessary to make the surveillance grid a working investment on my blog. Check it out here: http://www.inhardfocus.com
It’s really peculiar to *not* have the SFPD manage the camera project. I’ve seen a few small projects managed by non-profits (i.e. Columbus OH – project managed by Columbus Compact http://www.colscompact.com/camera.html), but any large scale project utilizing public funding is typically managed by the local PD.
We have tried to explain to cities for years that actual application of video surveillance. Video surveillance is deployed as a forensic tool meaning that unless someone is watching it, they can only go back and see what happens. The reality is that the quality of the video is typically very poor for the distance. It doesn’t matter how many frames per second, it’s the fact that the quality is so bad, you can’t use the evidence in a court of law. If you upgrade the quality to something that is reasonable, you are looking at 6 frames per second and 15Mbps of bandwdith per camera. That makes wireless a tougher proposition.
That alternative is to deploy a video analytic system that can proactively alarm when something is wrong. That is far better than a forensic system. A live body can make a decision on how to act based on immediate video feedback.
There are other options that my company has designed also so agencies that are looking at video surveillance need to really research the options and think through what they want the surveillance for before deployment.
Whilst I agree the quality of much surveillance is pretty terrible, 15Mbps per camera is highly misleading. With H.264 encoding, you can run DVD quality full frame video in as little as 2-3Mbps per camera. That is entirely possible over wireless. The essential factor is to use wireless technology that has deterministic bandwidth, not bog standard Wi-Fi.
My company has been involved in the specification and deployment of tens of successful municipal wireless surveillance systems in the UK and Ireland. The largest schemes have over 100 cameras.
Video analytics have a role to play – they’re fine for intruder and traffic violation detection – but they do not today have the intelligence to detect suspicious behaviour reliably in a complex street scene.
Digital video surveillance can do a lot of things such as monitor a site, record a scene and relay an event which would definitely SUPPORT law enforcement agencies and public safely organizations perform their duties and responsibilities. The argument that such systems when deployed in a public setting MUST accomplish the near-impossible task of CRIME DETERRENCE is a myopic view. That’s the domain of law enforcement and public safety NOT video surveillance!
To rebut this argument, let me cite the use of digital video surveillance in the private sector. The reason why banks, retail shops, casinos, hotels and other businesses deploy such systems is NOT to deter crime but to HELP support security personnel accomplish their job function which is to secure their facilities. Crimes will still occur of course but these businesses will have at least the video footage to help them prosecute the criminals, review their business processes and possibly revise their workplace policies. In other words, the digital video surveillance system functions as a kind of visual feedback which will ultimately benefit the enterprise.
Civil libertarians have this Luddite tendency to blame technology for the failure of U.S. society to address inner-city poverty, crime and despair. But that’s a social problem NOT a technological issue!
As a detective I have seen many a wireless camera used in an attempt to prevent crime. Where it may not prevent crime, it often is helpful in the investigation following the crime. I brought what I learned to Airtight Video and put our heads together and came up with a camera that works with motions sensors so it could record more frames per second without using up more hard drive and invisible infrared, (infrared usually glows red and is easily seen), so as not to be seen at night. Making these adjustments did not improve the ability to stop crime but it became much more useful in the investigations following the crimes with better resolution and more frames per second to work with.