What could they have done with ATM crisis?

Many ATMs including Hong Kong locations have been found contaminated with pinhole cameras and illegal card readers over the past month. Credit Card fraud from various kinds of devices used for skimming information from the credit cards of unsuspecting customers not only results in millions of dollars of losses but it significantly affects customer satisfaction and their confidence in their banker.

Many methods have been tried to thwart attempts at skimming credit cards and of course the miscreants have developed techniques for overcoming them.

iOmniscient has developed a video based capability which can detect skimming devices (even those that are almost invisible to the human eye). Further the system can detect the person who placed the device and, if he is in a database of known criminals, the system can recognize him. The capability is incorporated in iQ-Bank, a comprehensive portfolio of products specially tailored for the banking industry.

iQ-Bank supports the needs of different stake holders in the bank with a single integrated system. It can count the traffic and manage queues. It can help the Occupational Health and Safety team respond if a customer has slipped and fallen, it can recognize known culprits, detect theft of assets and generally improve the security, safety and efficiency of the bank.

Click for more details on iQ-Bank.

In future thieves who want to place devices on ATMs will be easily recognized and caught.

 

ATM Skimming Device Detected

*Eddie Luk, 13/11/2012, Alert as camera crooks hit bank ATMs, The Standard

New Smart City Book launched

Having won some very large Smart City projects around the world (the largest of which involves over 20,000 cameras) iOmniscient has put together the knowledge it has gained into a new book called iQ-Smart City, Security-Safety-Service. This is the first publication focused on how to implement successful technologies for a Smart City. There is a particular focus on maximizing the return on investment demonstrating that a good design can reduce the cost of implementation by an order of magnitude.

iQ-Smart City, Security-Safety–Services discusses the various technologies that can be used to manage a City environment and discusses the processes that must be used to ensure a successful implementation.

Comprehensive Smart City Solution that includes video, sound and smell analytics

Traditionally video analysis was defined as technologies based on simple motion detection. License Plate Recognition (LPR or ANPR) and Face Recognition were offered by suppliers of biometric solutions for controlled and cooperative environments.

Today iOmniscient has brought together all these technologies so that they work together in uncontrolled non-cooperative environments of the type one confronts in a typical city surveillance situation. Further, the needs of customers have moved beyond just video analytics to analytics in general. To accommodate this iOmniscient recently introduced audio and smell analysis (e.g. for detecting gas leaks).

There is also much talk about PSIM systems (Physical Security Information Management) which are essentially systems that bring in information from a number of different types of sensors and then provide the operator with instructions for manual responses. iOmniscient has taken this concept to the next level with a patented ability to provide Automated Responses for given situations. Such systems can reduce the load on Operators who invariably become the bottlenecks during a crisis.

These concepts along with the ability to feed all the Meaningful Meta Data (MMD) generated by iOmniscient’s systems into Big Data engines for further analysis is what Smart Cities is all about. It goes beyond the narrow definitions of various individual technologies and allows all of them to operate together.

A city is a very complex environment and requires information from many different sources to function effectively. It has to coordinate traffic and ensure a safe environment for pedestrians. During special events the authorities have to manage crowds. Cities have railways and metros, sports stadiums, entertainment centers, car parks and airports and all of these have special requirements. But they also need to work together. If a criminal that was detected at the airport later turns up in a football stadium the authorities may need to be made aware of this.

More information will be available from www.iQSmartCity.com. Please check back in February.

Want a free copy of our Smart City book?
The book is available free in any quantity for iOmniscient’s signed Resellers, Systems Integrators and also for customers (as long as they cover the postage/ courier charges). For all others the book is available at US$20 +postage from book@iomniscient.com
Soft copies of the book can be downloaded free by iOmniscient’s signed up Resellers, Systems Integrators and by its customers. For others the cost is $20 per copy. Please contact book@iomniscient.com for a download authorization code.

How Smart is Artificial Intelligence?

In using video analysis, users sometimes expect their systems to operate perfectly. They feel that if a system cannot recognize a person with 100% accuracy it is useless. If a system cannot detect every person that falls down then it is useless.

Human systems and Artificial Intelligence based computer systems have their own individual strengths. For instance, with Face Recognition systems humans are very good at recognizing faces that they are familiar with (eg. close friends and relatives). For strangers, a human is unlikely to recognize anyone accurately when the database used for matching has more than 20 faces in it. Humans also tend to have short attention spans and can be easily distracted. This is where computers come into their own. They can operate continuously without being distracted. They don’t fall asleep or chit-chat with their neighbors or read the newspaper.

Recently, in a shopping mall, the owners were concerned as they had a list of around 500 known shoplifters. They would provide their images to the security team but their success at recognizing the culprits was negligible.

The iOmniscient Face Recognition system was put in to recognize these people. As the crowd was uncontrolled and was not necessarily cooperative, the accuracy of the system was around 70%. Was this bad? Compared with the alternative of attempting to recognize them using humans this was an extremely good result.

Humans however have much better intuitive judgement than computers. So when the computer was not sure, it could provide a short list of possible matches. With a very small cohort of faces to look at the human was much better able to make a final judgement on whether a person was on the list of suspicious characters.

In this situation humans and computers working together could achieve far better results than either working on their own.

In the same mall the customer had installed a Slip and Fall detection system. While it was very accurate in detecting people falling down when it was correctly used there were situations where people fell and they were not detected. This happened for instance where there were no cameras to do the detection or where the view was obscured by furniture or pillars. Nevertheless the system outperformed any human operating under similar circumstances and proved a great help in reducing the number of insurance claims made through the resulting speedy response.

So one should not expect computer based systems to be perfect. However they can often beat a human operating in a similar situation and therefore augment his/ her capability.
See our "Facial Recognition In a Crowd" live at Dubai INTERSEC on 15th-17th Jan 2013, 10am-7pm, booth# S3-122P.
for appointment.
info@iomniscient.com | www.iomniscient.com