Fast stereo matching using adaptive guided filtering

  • Authors:

  • Qingqing Yang;Pan Ji;Dongxiao Li;Shaojun Yao;Ming Zhang

  • Affiliations:

  • Institute of Information and Communication Engineering, Zhejiang University, Hangzhou 310027, China and Zhejiang Provincial Key Laboratory of Information Network Technology, Hangzhou 310027, China ...;Institute of Information and Communication Engineering, Zhejiang University, Hangzhou 310027, China and Zhejiang Provincial Key Laboratory of Information Network Technology, Hangzhou 310027, China;Institute of Information and Communication Engineering, Zhejiang University, Hangzhou 310027, China and Zhejiang Provincial Key Laboratory of Information Network Technology, Hangzhou 310027, China;Institute of Information and Communication Engineering, Zhejiang University, Hangzhou 310027, China and Zhejiang Provincial Key Laboratory of Information Network Technology, Hangzhou 310027, China;Institute of Information and Communication Engineering, Zhejiang University, Hangzhou 310027, China and Zhejiang Provincial Key Laboratory of Information Network Technology, Hangzhou 310027, China

  • Venue:
  • Image and Vision Computing
  • Year:
  • 2014

Quantified Score

Hi-index 0.00

Visualization

Abstract

Dense disparity map is required by many great 3D applications. In this paper, a novel stereo matching algorithm is presented. The main contributions of this work are three-fold. Firstly, a new cost-volume filtering method is proposed. A novel concept named ''two-level local adaptation'' is introduced to guide the proposed filtering approach. Secondly, a novel post-processing method is proposed to handle both occlusions and textureless regions. Thirdly, a parallel algorithm is proposed to efficiently calculate an integral image on GPU, and it accelerates the whole cost-volume filtering process. The overall stereo matching algorithm generates the state-of-the-art results. At the time of submission, it ranks the 10th among about 152 algorithms on the Middlebury stereo evaluation benchmark, and takes the 1st place in all local methods. By implementing the entire algorithm on the NVIDIA Tesla C2050 GPU, it can achieve over 30million disparity estimates per second (MDE/s).