Content-based image lookup with SURT algorithm
Abstract
Abstract: Content-Based Image Retrieval - CBIR (Content-Based Image Retrieval) includes concepts, purposes, models, components, functions and several content-based image retrieval systems. The image matching problem is a subproblem of the image query problem. Image matching is the matching of the disproportionate features of two images. In this paper, I mainly study the image matching method based on invariant features using SURF algorithm. SURF is an invariant detector and descriptor of points of interest with scale and rotation. This method is equivalent or even faster than the previous proposed methods with regard to repetition, specificity and robustness, it also makes computation and comparison faster.SURF achieves this by building on integrated images with multiple image folds through building on the strengths of leading detectors and descriptors (here using the phantom method). Hessian’s matrix to measure the detector and based on the distribution method for the descriptors).