Study and Analysis The Effect of Camera Calibration Methods on Quality of Stereo Images Rectification

Authors

  • batoul khalil tishreen university
  • Al Samawal Saleh Tishreen University

Abstract

Stereo calibration and rectification are considered to be one of the most important pre‐steps in computer vision, virtual reality and 3D reconstruction applications. Fisheye lenses, which has a large Field of View, have been receiving increasing attentions in recent years. But these  lenses add a high amount of distortion to the images. Calibration is also a must when rectifying stereo image pairs.  Rectification makes the correspondence done only along the x-axis of the rectified images.

Because only few  researches compared calibration algorithms and their impact on stereo image rectification. This paper presents a study and analysis  of  two state of the art  calibration algorithms which are  Scaramuzza algorithm which suggests  a generic camera model to calibrate all kinds of cameras. And Bouguet algorithm that is based on Bouguet’s work on calibration for pinhole camera model with the addition of some changes in distortions modeling. The performance evaluation of the algorithms have been done based on subjective measures using GIMP program. While the objective performance evaluation was done by studding mean square  rectification error parameter (MSRE). The results showed that Bouguet algorithm is more accurate and flexible, as it gave better results in stereo image rectification.

Published

2023-12-07

How to Cite

1.
خليل ب, السموءل صالح. Study and Analysis The Effect of Camera Calibration Methods on Quality of Stereo Images Rectification. Tuj-eng [Internet]. 2023Dec.7 [cited 2024May2];45(5):257-74. Available from: https://journal.tishreen.edu.sy/index.php/engscnc/article/view/14870