December 31, 2005
MDI Security Systems Launches VerifEye IVS(TM)
MDI is another vendor of smart camera software.
There new software, VerifEye Intelligent Video Security Software does the usual stuff:
The standard VerifEye IVS package offers the following features and capabilities:
-- People and Vehicle Tracking
-- Motion and Behavior Analysis
-- Facial Detection
-- Perimeter Monitoring
-- Intrusion Detection & Secure Area Monitoring
-- Unattended or Abandoned Object Detection
-- Asset Protection & Object Removal
-- Anti-Tailgating and Anti-Piggybacking
-- Object Counting & Statistical Reporting
-- Entry & Exit Directional Monitoring
-- Object Detection, Classification and Pattern Recognition
-- Automated Self-Learning
-- GPS and Port GIS Mapping
-- Multiple Camera Tracking
-- Dynamic Report Generation
-- Utilization of Infrared and Thermal Cameras
-- Video Stabilization
-- Automated 24/7 Video Monitoring with Real-Time Alerts
-- High-Tech Filters that Exclude Natural Background and Irrelevant Objects
Advanced intelligent add-on options for VerifEye IVS include:
MotionEye - A powerfully superior video motion detector. It is an excellent enhancement option to all Digital Video Recorders, security systems and sensors. MotionEye far outweighs the basic built-in features found in cameras and DVRs.
TrackerEye - A sophisticated technology that automatically tracks objects with Pan/Tilt/Zoom (PTZ) cameras and eliminates the manual control process by automatically following and tracking objects. In complex situations, TrackerEye makes intelligent decisions based on threat levels and distances; tracking targets until completely out of the camera's field of view.
IntrusionEye - A state-of-the-art video based solution to anti-tailgating, anti-piggybacking and anti-passback. IntrusionEye directly addresses the long-standing problem of "access bypass" faced by the security industry. It ensures compliance with access control procedures via video verification and can easily be added on to most systems, including MDI's award-winning SAFEnet and iTRUST integrated solutions.
Posted by rshah at 10:58 AM | Comments (0) | TrackBack
Video authentication
A PR on smart cameras that provide video authentication from IQinVision.
I have no idea whether there is a real problem and need for this technology, but it is interesting nonetheless.
IQinVision, market leader in high performance megapixel network cameras, smart IP cameras and network video recording systems, today announced the introduction of IQauthenticate on-camera image verification for all IQeye501 and IQeye300 Series Smart Network Cameras.
Legal use of digital images is limited because they are easily manipulated, often making the images inadmissible for prosecution. Many digital video recorders (DVRs) and network video recorders (NVRs) can detect tampered images after they have been processed but cannot verify whether the source of the video was authentic. Therefore, these widely-used video recorders can "authenticate" video that has been tampered with - IQauthenticate solves this problem.
Alan Brill, senior managing director of Technology Services for Kroll On-Track, believes on-camera image authentication will have a significant impact on the surveillance industry. "There is a real need to ensure that digital video images have not been tampered with," said Brill. "Without such controls, the use of digital video for prosecutions and other legal-related uses could be seriously compromised. With older film cameras you could ensure your video was real by tracking it from the time you removed the film from the camera. When the industry shifted to digital video this 'guarantee' disappeared."
IQeye cameras with IQauthenticate have a unique encrypted digital signature that identifies the IQeye that produced the image and detects if the image has been altered. Users simply submit an image to 'http://www.iqeye.com/IQauthenticate.html' for verification, and if even one pixel has been changed IQauthenticate will detect it.
Posted by rshah at 10:51 AM | Comments (0) | TrackBack
December 24, 2005
France copies UK on video security
From United Press International:
France copies UK on video securityPARIS, Dec. 22 (UPI) -- The French parliament Thursday followed Britain's example and authorized a huge increase in video security surveillance. The National Assembly approved a measure urged by Interior Minister Nicholas Sarkozy to dramatically boost video security surveillance throughout the country of 60 million people.
Sarkozy and French security chiefs were impressed by the rapid speed with which the British security services solved the case of who was responsible for the four suicide bombings in the London mass transit system on July 7 that killed 52 people and wounded more than 700. Surveillance cameras at London's King's Cross station played a central role in identifying the perpetrators.
There are more than 3 million video surveillance cameras already operating in Britain and less than 50,000 in France. The new law will open the way for a massive increase in video surveillance and the installation of video cameras in public places including airports, railway and bus stations, mosques, churches and synagogues.
Posted by rshah at 08:20 AM | Comments (0) | TrackBack
December 16, 2005
Congressional Hearings on Border Surveillance System
I ran across a press release on the congressional hearing concerning the border surveillance system. The hearing focused on the Inspector General's review of the effectiveness of border surveillance, remote assessment, and monitoring technology. The statement by the IG Skinner is here (which I refer to as a report), while a press release by the Committee on Homeland Security is here.
Update: The OIG report can be found at the DHS web site.
I am interested in this from the perspective of smart cameras and not going to discuss the procurement and management issues noted by the IG. I have only read the IG report, so I don't know what else happened in the hearings. But the IG report has some good information. For background, the agency is question is the Department of Homeland Security’s (DHS) Bureau of Customs and Border Protection (CBP), who detects illegal entry into the United States and the hearing was titled "Mismanagement of the Border Surveillance System and Lessons for the New Secure Border Initiative."
The report notes that there are over 11,000 sensors (seismic and magnetic) that provide remote detection capability. These sensors are along both our borders. There are 255 camera sites (both color and IR) with 168 sites that are currently incomplete. The system includes a tracking component to ensure every alert creates a ticket and the result of the agent's investigation.
The IG found that the system is not well integrated (especially when you consider what was promised). For example, there is no integration between cameras and sensors. You would expect when a sensor alert is triggered that a camera would then pan to the area. Instead, this means a camera operator has to manually move the camera to find the cause of the sensor alert.
The IG notes the grand claims for the system and lack of quantitative data on the system.
But the bottom line is that 90% of the alerts were false alarms: something other than illegal alien activity, such or animals. On the southwest border, only two percent of sensor alerts resulted in apprehensions; on the northern border, less than one percent of sensor alerts resulted in apprehensions.
The report notes that despite claims by the agency, there was considerable manpower devoted to responding to false alarms. The IG also points out that without mobile surveillance, it will be difficult to respond the changing patterns of illegal aliens. (They will just route around existing camera sites).
The take away points for me are:
1. Contractors oversell
2. Integration is important - smart camera systems work best when components can interact with each other without human intervention
3. Quantitative data is important to evaluate systems - No one wants to collect that data (because they think it will make them look bad), and as a result its difficult to gauge the effectiveness of these systems.
4. False alarms are time consuming. It seems to me there needs to be a feedback loop from the users to the designers once the system is implemented to try to reduce these false alarms.
Update:
A story by FCW on the hearings
A story by Washington Technology on the report
Posted by rshah at 08:50 PM | Comments (0) | TrackBack
December 13, 2005
Cameras with embedded smarts
From Security Park:
IMS Research predicts that the next generation of network cameras will have embedded video content analysis capability to detect various behaviours or events. The first products are forecast to hit the market within the next twelve months.
Posted by rshah at 04:01 PM | Comments (0) | TrackBack
December 11, 2005
Gunshot Cameras
A story on gunshot detection camera system (SENTRI) is in Government Technology. It covers the basics on how the system works and its creators. The best nugget comes out of the discussion on the chicago deployment and notes that it is currently in a pilot state.
"Pilot stage is really a technical term," said Bryan Baker, chief executive officer of Safety Dynamics, which produces the gun recognition technology. "When they define something as a pilot, that means there's still a certain confidentiality about information. Once it gets reclassified as production and not pilot, then all the information becomes nonconfidential."
So maybe once its out of pilot state, the rest of us will get more data on how well these systems are operating. The article also suggests what the next step maybe in these systems:
Several companies are creating a video analysis component that would recognize a shooting scene -- the position someone would be in when holding a gun, someone lying on the ground or a group of people running. "They can digitize what somebody holding a gun would look like," Baker said. "Then they can lock in and follow him. When you put all these things together, it can be successful in protecting parameters. You make it increasingly difficult for an intruder to escape notice."
Posted by rshah at 07:40 PM | Comments (0) | TrackBack
Cruise Ships and Cameras
From the Herald.com:
An article that discusses in passing surveillance cameras on cruise ships. Part of the article states:
Another lawsuit was filed against Carnival Cruise Lines last month in Miami by the family of a Wisconsin woman who disappeared from a ship a year ago and is presumed dead.
The suit says Carnival could have prevented her death if its security officers had been paying attention to a surveillance camera near the area where her purse and other belongings were found. The camera was covered with a piece of cloth, which should have alerted security officers, the suit contends.
So a camera gets covered up and the operator doesn't realize it. I don't know the setup, but it makes you wonder if a smart system would have helped in this case. Were people watching the systems and missed something? Or did people just ignore the surveillance system?
Posted by rshah at 07:34 PM | Comments (0) | TrackBack
December 07, 2005
Facial Recognition for the Masses
Wired has a story on Riya, which is a new photo-tagging service that uses facial-recognition technology to identify the people in your pictures. The goal is to quickly identify people in everyday pictures. Its a nice example of finding new uses for facial recognition technology.
Posted by rshah at 07:10 AM | Comments (0) | TrackBack
December 06, 2005
Definitive Paper on Smart Cameras
For those looking for a meaty (and academic) paper on smart cameras, we finally have one. It is a tad heavy on Velastin's research, but its the best general paper on smart cameras that I have seen. It also cites over a 100 other papers. Here is the abstract and a link to a posted pdf version:
This survey describes the current state-of-the-art in the development of automated visual surveillance systems so as to provide researchers in the field with a summary of progress achieved to date and to identify areas where further research is needed. The ability to recognise objects and humans, to describe their actions and interactions from information acquired by sensors is essential
for automated visual surveillance. The increasing need for intelligent visual surveillance in commercial, law enforcement and military applications makes automated visual surveillance systems one of the main current application domains in computer vision. The emphasis of this review is on discussion of the creation of intelligent distributed automated surveillance systems. The survey concludes with a discussion of possible future directions.
Citation:
M. Valera and S.A. Velastin, Intelligent distributed surveillance systems: a review, IEE Proc.-Vis. Image Signal Process., Vol. 152, No. 2, April 2005, p.192-204.
Posted by rshah at 04:02 PM | Comments (0) | TrackBack
Indentix's Facial Recognition System
Facial recognition funding is still ongoing despite EPIC's disapproval.
The latest is an order from the U.S. Department of State for its biometric program (ABIS) to Indentix. Identix was selected through a competitive bidding process and detailed technical evaluation. According to Security Park's rehashing of the PR, Identix performed well:
In the first six months of the original terms of the contract, Identix deployed three fully operational systems in support of the Visa Office, including an ABIS system with more than 40 million enrolled images and the e-Diversity Visa system. Within the first few weeks of operation of the e-Diversity Visa system, DOS credited Identix' ABIS with helping to identify more than 5,000 individuals attempting to enter the U.S. under fraudulent conditions.
In fact, at one site, there was a claim that Identix was able to successfully differentiate between identical twins! No idea on the veracity of this claim (or the limitations), but it is interesting. According to COTS Journal, Identix's solution is noteworthy because it runs on everyday computers. The figure below comes from their article and illustrates how Identix's facial recognition works.
While I share EPIC's mistrust of facial recognition, my distrust is based on past performance and the notion that facial recognition is the "silver bullet". Instead, I see facial recognition as useful under carefully controlled situations with oversight. It looks like its working for the Department of State, lets hope as time goes on, we can find out more about the effectiveness of this technology.
Posted by rshah at 03:38 PM | Comments (0) | TrackBack