q-Recursive Method to Improve Accuracy and Time Taken by Content Based Image Retrieval
Nidhi Bansal · International journal of engineering and future technology · 2018
Content Based Image Retrieval is an active and fast advancing research area for manipulating large amount of image databases. It is based on extracting and comparing the visual attributes of the images. Examples of visual attributes are color, texture, shape, and motion parameters. In order to extract features of an image, various feature extraction methods are available. One of them is moment description. The Zernike Moment Descriptor is a moment based Shape Descriptor. In this paper, we proposed that the process of Content Based Image Retrieval can be made fast and more accurate by using the q-recursive method to compute Zernike Moments. The method retrieves most relevant images according to the similarity measure calculated between features of the query image and images of the image database.