Object segmentation based on disparity estimation

ABSTRACT

Object segmentation plays an important role in multi-view video analysis. In this paper, we present a new object segmentation method for multi-view video in which only the disparity is used for segmentation and the motion estimation is neglected. Firstly, a modified locally adaptive support-weight approach is proposed for disparity estimation. Then, segmentation is realized by mean-shift algorithm. The experimental results show that proposed method could segment the semantically meaningful objects from complex background with high precision.

REFERENCES

• 1: E. Francois ,B. Chupeau, 1997. Depth-based segmentation. IEEE Trans. on CSVT, Vo1.7 (Feb.1997), 237--239.
• 2: E.Izquierdo, S. Kruse, 1998. Image analysis for 3D modeling, rendering, and virtual view generation. Computer Vision Image Understanding, Vo1.71, No.2 (1998), 231--253.
• 3: Yu-lin Chang, Chih-ying Fang, etal, 2007. Depth Map Generation for 2D-to-3D Conversion by Short-Term Motion Assisted Color Segmentation. IEEE International Conference on Multimedia and Expo, (July 2007), 1958--1961.
• 4: Cigla C. Aydin Alatan A, Depth Assisted Object Segmentation in Multi-View Video. 3DTV Conference: The True Vision -- Capture, Transmission and Display of 3D Video (May 2008), 185--188.
• 5: Thakoor N, Jean G, Devarajan V, Multihypothesis Prior for Segmentation of Stereo Disparity. IEEE Signal Processing Letters, Volume 15, (2008), 613--616.
• 6: Kuk-Jin Yoon , In So Kweon, Adaptive Support-Weight Approach for Correspondence Search, IEEE Transactions on Pattern Analysis and Machine Intelligence, v.28 n.4, p.650, April 2006 [doi>10.1109/TPAMI.2006.70]
• 7: Dorin Comaniciu , Peter Meer, Mean Shift: A Robust Approach Toward Feature Space Analysis, IEEE Transactions on Pattern Analysis and Machine Intelligence, v.24 n.5, p.603-619, May 2002 [doi>10.1109/34.1000236]
• 8: Taguchi Y, Wilburn B, Zitnick C L, Stereo reconstruction with mixed pixels using adaptive over-segmentation. IEEE Conference on Computer Vision and Pattern Recognition, (June 2008), 1--8.
• 9: Ping An, Chaohui Lu, Zhaoyang Zhang, Object segmentation using stereo images. International Conference on Communications, Circuits and Systems 2004, Volume 1 (June 2004), 534--538.

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
        I.2.10 Vision and Scene Understanding
            Subjects: 3D/stereo scene analysis

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
        I.2.10 Vision and Scene Understanding
            Subjects: Intensity, color, photometry, and thresholding

General Terms:
Algorithms, Experimentation, Performance

Keywords:
disparity estimation, locally adaptive support-weight approach, object segmentation