Comparative Analysis of Homography Technique based on RANSAC and Least Square Method
Geetanjali Babbar, Rohit Bajaj · 2023
Color homography is the main component of geometric approaches used in computer vision and is utilized in camera calibration, stereoscopic vision, 3D views, and many other applications. It links the visual components of two images with the help of the Root Polynomial and Least Square Method. These methods are helpful when there is no change in the viewing conditions of images. In this paper, we propose that colors across a change in viewing condition can also be related by homography such as changing light color, shading, camera, etc. This study presents an integrated framework that describes the various phases of the homography estimation method. This study also provides an overview of two approaches i.e. direct and feature-based used for homography estimation. In this research paper, a comparative analysis of Root Polynomial, Least Square, and other techniques are performed on the different datasets with parameters of Precision, Recall, F-measure, Mean, and Median. These parameters are used to determine the simulated results.