Smart Cameras Blog

[From Video analytics: Video analytics experts point at computer vision as the key to accuracy in intelligent video surveillance]

A PR laden article that has a few useful insights into smart cameras. It notes limitations of smart cameras and then discusses their applications for retail stores, marketing research, and advertising.

Limitations

However, computer vision has yet to achieve the level of intelligence critical to video analytics systems. “I think one of the limitations of currently available video analytics systems is their accuracy level. Most applications are not 100 percent accurate because computers still lack the visual intelligence that humans possess,” noted Michelle Tecson, sales and marketing executive associate at Neugent. In people counting, for example, computers currently count blobs—blobs of people coming in, whether two people entering together side by side or one large person with a shopping cart. Both will be counted as one. When a person is standing directly behind another person sitting down, with his back on the sitting person, the computer will recognize them as one instead of two.

Applications:

Retail stores use video analytics to identify hotspots within a store. The term “hotspot” was coined by weather channels that show typhoon areas in color codes. In the retail arena, hotspots are places within the store that are most frequented by customers. By accumulating data on customers’ movements within the store over a period of time, the storeowner can create a color-coded map identifying hotspots. Based on the identified hotspots, the storeowner can plan the future layout of the store or the positioning of merchandise to boost visibility of items and increase profits. . . .

Video analytics in market research and advertising is a slightly more sophisticated system, but the benefits are real and tangible. Mariano observed that in some applications, face recognition technology is involved. “It’s a fairly new DVR application. Some companies in the US, such as Brickstream and Shoppertrak, are installing DVR solutions to analyze customer behavior in such places as banks, retail stores, grocery stores or gasoline stations,” he said. Mariano cited one project his former company had with McDonald’s in Philadelphia in the US, which involved counting cars going through drive-thru lanes, and detecting and classifying the gender of drivers based on their facial characteristics. They discovered that the DVR solution was 70 percent accurate, which was not bad for a machine. . . .

Advertising agencies have also introduced some video analytic solutions that measure the impact of an ad to a passerby or bystander. It detects a face (if it’s male or female) as well as the face’s movement, whether it is looking at the ad on the wall and for how long, or looking away from it.

rshah Applications

[From China's All-Seeing Eye : Rolling Stone]

A nice article on surveillance in China, with a bit on the role of smart cameras, facial recognition technology, and the role of US companies.

Now, as China prepares to showcase its economic advances during the upcoming Olympics in Beijing, Shenzhen is once again serving as a laboratory, a testing ground for the next phase of this vast social experiment. Over the past two years, some 200,000 surveillance cameras have been installed throughout the city. Many are in public spaces, disguised as lampposts. The closed-circuit TV cameras will soon be connected to a single, nationwide network, an all-seeing system that will be capable of tracking and identifying anyone who comes within its range — a project driven in part by U.S. technology and investment. Over the next three years, Chinese security executives predict they will install as many as 2 million CCTVs in Shenzhen, which would make it the most watched city in the world. (Security-crazy London boasts only half a million surveillance cameras.)

The security cameras are just one part of a much broader high-tech surveillance and censorship program known in China as “Golden Shield.” The end goal is to use the latest people-tracking technology — thoughtfully supplied by American giants like IBM, Honeywell and General Electric — to create an airtight consumer cocoon: a place where Visa cards, Adidas sneakers, China Mobile cellphones, McDonald’s Happy Meals, Tsingtao beer and UPS delivery (to name just a few of the official sponsors of the Beijing Olympics) can be enjoyed under the unblinking eye of the state, without the threat of democracy breaking out. With political unrest on the rise across China, the government hopes to use the surveillance shield to identify and counteract dissent before it explodes into a mass movement like the one that grabbed the world’s attention at Tiananmen Square.

Regarding facial recognition:

In Guangzhou, an hour and a half by train from Shenzhen, Yao Ruoguang is preparing for a major test of his own. “It’s called the 10-million-faces test,” he tells me. Yao is managing director of Pixel Solutions, a Chinese company that specializes in producing the new high-tech national ID cards, as well as selling facial-recognition software to businesses and government agencies. The test, the first phase of which is only weeks away, is being staged by the Ministry of Public Security in Beijing. The idea is to measure the effectiveness of face-recognition software in identifying police suspects. Participants will be given a series of photos, taken in a variety of situations. Their task will be to match the images to other photos of the same people in the government’s massive database. Several biometrics companies, including Yao’s, have been invited to compete. “We have to be able to match a face in a 10 million database in one second,” Yao tells me. “We are preparing for that now.”

The companies that score well will be first in line for lucrative government contracts to integrate face-recognition software into Golden Shield, using it to check for ID fraud and to discover the identities of suspects caught on surveillance cameras. Yao says the technology is almost there: “It will happen next year.” . . . Yao denies that a primary use of the technology he is selling is to hunt down political activists. “Ninety-five percent,” he insists, “is just for regular safety.”

Interesting is the lack of recognition from its partner company, based in the US:

L-1’s enthusiasm is hardly surprising: If Yao impresses the Ministry of Public Security with the company’s ability to identify criminals, L-1 will have cracked the largest potential market for biometrics in the world. But here’s the catch: As proud as Yao is to be L-1’s Chinese licensee, L-1 appears to be distinctly less proud of its association with Yao. . . . The company’s reticence to publicize its activities in China could have something to do with the fact that the relationship between Yao and L-1 may well be illegal under U.S. law. After the Chinese government sent tanks into Tiananmen Square in 1989, Congress passed legislation barring U.S. companies from selling any products in China that have to do with “crime control or detection instruments or equipment.”

But they are not the only US companies involved with China:

As The New York Times recently reported, aiding and abetting Beijing has become an investment boom for U.S. companies. Honeywell is working with Chinese police to “set up an elaborate computer monitoring system to analyze feeds from indoor and outdoor cameras in one of Beijing’s most populated districts.” General Electric is providing Beijing police with a security system that controls “thousands of video cameras simultaneously, and automatically alerts them to suspicious or fast-moving objects, like people running.” IBM, meanwhile, is installing its “Smart Surveillance System” in the capital, another system for linking video cameras and scanning for trouble, while United Technologies is in Guangzhou, helping to customize a “2,000-camera network in a single large neighborhood, the first step toward a citywide network of 250,000 cameras to be installed before the Asian Games in 2010.” By next year, the Chinese internal-security market will be worth an estimated $33 billion — around the same amount Congress has allocated for reconstructing Iraq.

rshah Other Cities

[From CCTV boom has failed to slash crime, say police | UK news | The Guardian], See also, this story from the Independent and comments on Schneier’s blog.

The article discusses a why the cameras haven’t worked to reduce crime and what the UK is doing to fix the problem. Here are some snippets:

Only 3% of street robberies in London were solved using CCTV images, despite the fact that Britain has more security cameras than any other country in Europe. The warning comes from the head of the Visual Images, Identifications and Detections Office (Viido) at New Scotland Yard as the force launches a series of initiatives to try to boost conviction rates using CCTV evidence. They include:

· A new database of images which is expected to use technology developed by the sports advertising industry to track and identify offenders.

· Putting images of suspects in muggings, rape and robbery cases out on the internet from next month.

· Building a national CCTV database, incorporating pictures of convicted offenders as well as unidentified suspects. The plans for this have been drawn up, but are on hold while the technology required to carry out automated searches is refined.

rshah Policy

[From City picks gunshot detection system - NJ.com]

Newark has chosen Shotspotter for a gunshot detection system over competing systems from Safety Dynamics (SENTRI) and Planning Systems (SECURES).

Gunshot detectors will be mounted on structures over a 7-square-mile area. When triggered, Shotspotter sensors will provide police with the location, number of shooters, and number of shots fired. . . . Shotspotter’s Gunshot Location System has also been installed in 27 cities around the country, including Washington, D.C., and East Orange. Newark’s neighboring municipality first introduced gunshot detectors made by Planning Systems Inc. in 2005, and deployed a Shotspotter system in 2006, according to Detective Andrew DiElmo of the East Orange Police Department. Both systems are in operation in East Orange.

Gunshot detectors are viewed as a “useful tool” for investigative purposes, and the East Orange police also considers it a crime deterrent, DiElmo said. However, there have been issues about gunshots that go undetected by sensors. “Does it work 100 percent of the time? No,” said DiElmo. “We have had incidents where the sensors haven’t alerted us.”

rshah Gunshot Detection, Other Cities

[From The Key Challenge for Video Analytics : IP Video Market Info]

John Honovich points out the importance of false positives. A false positive arises when a smart camera system warns operators of a situation, when the situation is normal. This problem arises when cameras move out of laboratories into real world conditions where weather and external stimuli, such as birds, make video analytics difficult.

Here are some interesting snippets:

Making sure you only trigger when someone is violating is hard because there are so many factors that might set off the analytic besides a valid adversary. To a computer, rain, dust and snow can all generate a similar form to a human being. Quick changes in light or the movement of water (waves) can also generate such forms. A camera that shakes because of the wind or issues with the mounting or installation also can trigger such alerts. The hard part in such analytics is to make sure that these alerts can be eliminated. This is a key metric in testing and differentiating between analytics.

False positives drive up the cost of systems. . . . While it is better operationally to centrally manage alerts, if the system generates dozens or hundreds of false alerts per day, the costs can become prohibitive. Let’s say an ‘intelligent’ camera generates 5 false alerts a day at the cost of $1 per alert (the unit cost to pay a monitor to assess). That’s $5 per day, over $1,800 per year and about $10,000 for a projected 5 year lifecycle. If you dozens or hundreds of cameras, this hidden operational cost can be in the millions. And this is not theoretical. This is the feedback you will hear time and again from real world deployments. It’s widely accepted that this is improving but it is still the major factor in duing your due diligence in analytics.

rshah Applications

[From D.C. Forging Surveillance Network - washingtonpost.com]

The network is expanding by 5000 cameras. It will include 3 to 5 operators watching images in real time. Analytic software is expected to be added by the end of the year. The cameras come from existing networks:

In its start-up phase, the system will include the public schools, the D.C. Housing Authority, the Office of Property Management and the Transportation Department. By year’s end, it will expand to homeland security and the departments of Parks and Recreation, Corrections, Health and Fire and Emergency Medical Services. The schools have the largest number of cameras, about 3,600.

Right now DC is still looking for funding and privacy advocates are not happy that the department is rushing ahead without applying prior safeguards. Until now, DC had been the model for the best privacy safeguards. See the article for more or read this nice opinion piece by John Podesta and William Sessions.

rshah Other Cities