Nonmetric calibration of camera lens distortion: differential methods and robust estimation

Electr. Eng. Dept., Assiut Univ., Egypt
IEEE Transactions on Image Processing (Impact Factor: 3.2). 09/2005; DOI: 10.1109/TIP.2005.846025
Source: IEEE Xplore

ABSTRACT This paper addresses the problem of calibrating camera lens distortion, which can be significant in medium to wide angle lenses. Our approach is based on the analysis of distorted images of straight lines. We derive new distortion measures that can be optimized using nonlinear search techniques to find the best distortion parameters that straighten these lines. Unlike the other existing approaches, we also provide fast, closed-form solutions to the distortion coefficients. We prove that including both the distortion center and the decentering coefficients in the nonlinear optimization step may lead to instability of the estimation algorithm. Our approach provides a way to get around this, and, at the same time, it reduces the search space of the calibration problem without sacrificing the accuracy and produces more stable and noise-robust results. In addition, while almost all existing nonmetric distortion calibration methods needs user involvement in one form or another, we present a robust approach to distortion calibration based on the least-median-of-squares estimator. Our approach is, thus, able to proceed in a fully automatic manner while being less sensitive to erroneous input data such as image curves that are mistakenly considered projections of three-dimensional linear segments. Experiments to evaluate the performance of this approach on synthetic and real data are reported.

  • [Show abstract] [Hide abstract]
    ABSTRACT: Camera calibration is a fundamental and important step in many machine vision applications. For some practical situations, computing camera parameters from merely a single image is becoming increasingly feasible and significant. However, the existing single view based calibration methods have various disadvantages such as ignoring lens distortion, requiring some prior knowledge or special calibration environment, and so on. To address these issues, we propose a line-based camera calibration method with lens distortion correction from a single image using three squares with unknown length. Initially, the radial distortion coefficients are obtained through a non-linear optimization process which is isolated from the pin-hole model calibration, and the detected distorted lines of all the squares are corrected simultaneously. Subsequently, the corresponding lines used for homography estimation are normalized to avoid the specific unstable case, and the intrinsic parameters are calculated from the sufficient restrictions provided by vectors of homography matrix. To evaluate the performance of the proposed method, both simulative and real experiments have been carried out and the results show that the proposed method is robust under general conditions and it achieves comparable measurement accuracy in contrast with the traditional multiple view based calibration method using 2D chessboard target.
    Optics and Lasers in Engineering 12/2013; 51(12):1332-1343. · 1.92 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Fish-eye cameras are widely used on many occasions due to their ultrawide field of view (about 180°). In this paper, we present a high-precision two-step calibration method to calibrate fish-eye cameras. The two steps are the global polynomial projection model fitting and local line-fitting calibration optimization. In the first step, we obtain the projection model of the fish-eye camera and apply a quartic polynomial to fit the projection model over the entire image. In the second step, the fish-eye image is partitioned into several sections and line fitting is adopted in each section in order to further reduce the residual error of the first calibration step. Experiments show that the new method is able to correct the distortion of the real scene image well. In addition, its average reprojection error is 0.15 pixel better than 0.40 pixel of the general projection model described. The reason that higher calibration precision is obtained is that this method not only considers the global projection model of the fish-eye camera but also considers the local characteristics, such as small tangential distortion and asymmetry.
    Applied Optics 03/2013; 52(7):C37-42. · 1.69 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: It is advantageous but difficult to automatically and accurately estimate parameters of lens models from single images, particularly in internet image processing and thermal image processing. As not all objects have straight edges, it is natural that some edges should be straight and some should be curved in undistorted images. Therefore, the key problem is how to find curved lines in original images that should be straight in undistorted images. In this paper, a novel algorithm is proposed to solve this key problem by automatically selecting reliable line segments using high-order Hough transform energies. Experimental results show that the proposed algorithm is able to detect reliable lines and produce quality results for both indoor and outdoor images taken by different cameras using wide-angle and fisheye lenses as well as zoom lenses.
    Applications of Computer Vision (WACV), 2013 IEEE Workshop on; 01/2013