Image Retrieval using Color and Texture features

“

Due to the rapid increase of image capturing equipment and digitalization throughout the world, millions of images are being created in every second. Among these huge number of images, some images are needed to available for the user for their future purpose. The retrieval of images is more complex than image capturing and storage. For making the captured images available to the user, various methods have been introduced by the researchers. Text-based image retrieval is the traditional image searching method where all the images are needed to be annotated manually. It is so complex because a huge labor is needed to annotate all the images, also time-consuming. To eliminate this complexity, Content-based image retrieval method has been introduced and for last three decades, it is the hot topic in the world of information technology. Here we have introduced a new method that comprises the both color and texture features for retrieving images from the vast database with respect a query image providing with better accuracy within a little time compared to the existing CBIR method. In the next research in this field Local Tetra Pattern, Contourlet transform will be used instead of wavelet transform for texture feature extraction. I believe, this technique will provide accuracy more than 87.5%.

”