Affine Invariant-Based Classification of Inliers and Outliers for Image Matching

Abstract

This paper presents a new approach to classify tentative feature matches as inliers or outliers during wide baseline image matching. After typical feature matching algorithms are run and tentative matches are created, our approach is used to classify matches as inliers or outliers to a transformation model. The approach uses the affine invariant property that ratios of areas of shapes are constant under an affine transformation. Thus, by randomly sampling corresponding shapes in the image pair we can generate a histogram of ratios of areas. The matches that contribute to the maximum histogram value are then candidate inliers. The candidate inliers are then filtered to remove any with a frequency below the noise level in the histogram. The resulting set of inliers are used to generate a very accurate transformation model between the images. In our experiments we show similar accuracy to RANSAC and an order of magnitude efficiency increase using this affine invariant-based approach.