Computer Vision: A Talk by Prof. A. K. Jain

The IGNCA Asian Conference New Delhi, 15-29 November 1996.IGNCA, 4th July 1996 Prof. A.K. Jain, an eminent name in image processing and computer vision, gave a talk on his work underway at Michigan State University, College of Engineering, Computer Science Department, Michigan.

An interesting application serves as the building block of ‘virtual museums’ and reconstructs a 3-D model of Buddha bust from nine images which are displayed by an angle to capture the complete information of the bust. With the Internet, virtual museums are now a reality.

Another challenging project involves old silver tankards. These are old, precious artifacts, often several hundred years old. The fundamental problem in dealing with them is their authentication. As in several other applications across the globe, authentication is one of the primary uses for which image analysis and computer aided verfication is employed. The tankards have the mark of the artist and the city of its origin. Over a time, these get worn out and only a blurred image is left. The application involves digitizing the mark, enhancing it and then extracting its primary features for verification.

An application of image processing relevant for both the corporate and scholarly communities is the ability of a system to read, i.e. optical character recognition (OCR). Document imaging has emerged as an important and rapidly developing area. The success rate of several OCR systems often exceeds 99%. A high success rate is required for this system to be meaningful. If the rate is one lesser than 98%, it would often be cheaper to re-enter the document.

One common application is in the field of quality assurance, the imaging and subsequent analysis of the objects. It is applied for a fast and reliable quality check.

A unique application of computer vision is one involving image compression. A company has attempted to put the image of a credit card holder on the card itself, leading to higher security.

There are several other applications of computer vision in the areas of biometrics, industrial inspection and quality assurance. These growing and unusual applications make computer vision one of the grand challenges in the coming years.

 

 

[ Newsletter | List of Newsletter ]