Camera lens radial distortion correction using two-view projective invariants

Abstract

This Letter proposes a new method for automatically removing radial lens distortion from image feature point correspondences of two views. Based on the projective geometric relationship between two views of a planar scene, we have derived a system of algebraic equations that relates the invariants to the distortion parameters to be found. We then propose a noniterative procedure to solve the equations system, and a kernel-voting scheme to select the best root. Being a noniterative approach, our method overcomes many problems with the conventional iterative approach. It also largely decouples the estimation of the distortion from the estimation of other camera parameters and, therefore, delivers more reliable results. Experiments on both synthetic data and real images have provided satisfactory results.