Recognizes and Searches Visual Patterns

The famous search engines like GOOGLE work well with the text based searches. In this method the text is searched on the basis of keywords. But this method is not very effective in searching the images.

A technique has developed by Mirela Tanase (A Dutch Researcher) to make the search engines work like the human eye and find the images. In this method the object is divided into the different parts. Then similar objects are searched for each part. She also made a search engine which can find the images using these object parts. This search engine is also called the “part based” search engine. This method is an enhancement of the MPEG7 method which is used to search the contours. The part based search engine can give the 70% accurate results as compared to the MPEG7 method.

For example if you are going to search for a car, then in this method search will be done for every part of the car like wheels, doors, mirrors etc. The results will be the similar objects found. This method increases the chances of effective searching.