|
Introduction
September 11th has changed the way we live our life forever. There is consensus that availability of advanced technology for surveillance and intelligence would have prevented the tragedy.
What if intelligent biometrics systems were installed to help Identify terrorists before they could board the planes?
What if we had autonomous control of airplanes from ground which could have prevented the aircrafts from actually crashing into the World Trade Center? and the list goes on. One technology hopes to answer at least some of these burning questions.
Computer vision or the technology that helps computers to see has made giant strides in the past few years.
It has been an active research topic in the research community for the past 5 decades. The US defense has long used computer vision to develop UAVs (Unmanned Aerial Vehicles) and cruise missiles. High cost of computing and complex engineering issues had prevented it from becoming a commercial technology. With the $ per GHz of computing crashing thru the roof, computer vision algorithms can now be run on regular desktop.
Understanding computer vision usually requires attending a semester long boot camp at a large university. Skipping the boot camp details and jumping right into the Heart of the matter, computer vision is a subfield of study in the general area of artificial intelligence and its goal is to make computers smarter by helping them to "see" and make intelligent decisions through various heuristics.
There are various problems that computer vision addresses like, human computer interaction, biometrics, security, defense, market research, etc. But post 9/11, face recognition systems for identifying terrorists and criminals seems to be the hottest application. Various companies like Visionics and Viisage have already deployed commercial face recognition systems in Tampa downtown and Boston Logan International Airport. But the accuracy and performance of these systems are definitely questionable. In fact, ACLU (American Civil Liberties Union) posted a report blasting the face recognition technology being used in Tampa, Florida.
Though the performance and accuracy of computer vision systems will be paramount for their success in the mainstream, thanks to the huge academic research backing and increasing support by commercial businesses, CV (Computer Vision) hopes to do just that in the coming decade. And the applications will not be limited to just biometrics. The Artificial Intelligence concepts involved in computer vision will help it to do a lot of things like enable Intelligent human computer interfaces, perform human body tracking, do Augumented Reality, provide Infotainment, etc.
A visit to the unofficial computer vision homepage shows the growing interest in the field. Also, Intel has created the OpenCV initiative with source code available at Source Forge. Microsoft, yes Microsoft also is contributing to the vision community by releasing the Vision SDK (of course its not open source) but the libraries are free to use.
>> How Does A Computer Vision System Work?
|
