In this paper, we are proposing an algorithm for a content-based method for retrieving images. In contrast to many previous computer-aided methods, which require intensive and accurate human intervention, this method needs only the user to provide a image database of the similar content as the query image.
For input, the method requires a true color source image database and a target image. The source and target image are both transformed into a perceptually de-correlated color space. In this color space a best matching source pixel is determined for each pixel of the target image. The matching criterion uses the first order statistics of the luminance distribution in a small window around the source and target pixels. Along with the luminance distribution, high pass filter is used to improve texture information in L-channel of the source and target images and edge detector is used in order to detect the edges of the source and target images. The only requirement of the method is that the compositions of the source and target scenes resemble each other.
Cite this article:
Umesh Agrawal, Yogesh Rathore , Achin Tripathi. An Efficient Method for Image Matching and Retrieval. Research J. Engineering and Tech. 1(2): Oct. - Dec.2010 page 74-78.
Umesh Agrawal, Yogesh Rathore , Achin Tripathi. An Efficient Method for Image Matching and Retrieval. Research J. Engineering and Tech. 1(2): Oct. - Dec.2010 page 74-78. Available on: https://ijersonline.org/AbstractView.aspx?PID=2010-1-2-5