Depth-Preserving Warping for Stereo Image Retargeting

Bing Li Lingyu Duan Chia-Wen Lin Tiejun Huang Wen Gao
bing.li@vipl.ict.ac.cn lingyu@pku.edu.cn cwlin@ee.nthu.edu.tw tjhuang@pku.edu.cn wgao@pku.edu.cn

 

 

Abstract
The popularity of stereo images and various display devices poses the need of stereo image retargeting techniques. Existing warping-based retargeting methods can well preserve the shapes of salient objects in a retargeted stereo image pair. Nevertheless, these methods often incur depth distortion, since they attempt to preserve depth by maintaining the disparity of a set of sparse correspondences, rather than directly controlling the warping. In this paper, by considering how to directly control the warping functions, we propose a warping-based approach to stereo image retargeting which can simultaneously preserves the shapes of salient objects and the depths of 3D scenes. We first characterize the depth distortion in terms of warping functions to investigate the impact of a warping function on depth distortion. Based on the depth distortion model, we then exploit binocular visual characteristics of stereo images to derive region-based depth-preserving constraints which directly control the warping functions so as to faithfully preserve the depths of 3D scenes. Third, with the region-based depth-preserving constraints, we present a novel warping-based stereo image retargeting framework. Since the depth-preserving constraints are derived without taking into account shape preservation, we relax the depth-preserving constraints so as to reach a good tradeoff between shape preservation and depth preservation. Finally, we propose a quadbased implementation of the proposed framework. Our results demonstrate the efficacy of our method in both depth and shape preservation for stereo image retargeting.

Experimental Results
  1. Effectiveness of Depth-Persevering Constraints
  • Quantitative results
Enlarging the image width by 10%


Shrinking the image width by 10%

  • Visualized Results on Baby1
Baby1

  1. Comparisons of Stereo Retargeting Methods

Snowman

Aloe

People

Man

Baby1

Diana

Car


Download