Skip to main content
SpringerLink
Log in

Disparity refinement based on least square support vector machine for stereo matching

  • Original Paper
  • Published:
Signal, Image and Video Processing Aims and scope Submit manuscript

Abstract

Disparity refinement is a crucial step in obtaining accurate disparity for stereo matching method. Outlier disparity value still exists in some areas (such as feeble texture and discontinuous regions), even the advanced stereo matching algorithm based on deep learning. To address this issue, a novel disparity refinement method based on the least square support vector machine (LSSVM) is proposed. In this method, the least square support vector machine model is first applied to every horizontal line of the obtained initial disparity map to model the disparity values, corresponding image color values, and coordinates of pixels. According to corresponding feature, the predicted disparity value of each pixel is calculated by this regression model. Subsequently, the outliers are detected and removed based on the residual between the real and predicted disparity value for obtaining more accurate initial disparity map. Then, along a horizontal line of the disparity map, the LSSVM with different parameters is applied to train the valid disparity values and its feature for obtaining the trained regression model. Finally, the invalid disparity values are redefined by the trained regression model. Experimental results demonstrate that the proposed method shows a better performance compared with current some disparity refinement methods. When the proposed algorithm is implemented on the disparity map of the deep learning method, the error rate has decreased and the maximum decline rate is 4.3 and 3.0 in nonocc and all regions, respectively.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7

Similar content being viewed by others

References

  1. H, K., A, H.: 3d scene reconstruction from multiple spherical stereo pair. Int. J. Comput. Vis. 104(1), 94–116 (2013)

    Article  MathSciNet  Google Scholar 

  2. Jia, X., Chen, W., Liang, Z.: Bidirectional stereo matching network with double cost volumes. IEEE Access 49, 1–9 (2021)

    Google Scholar 

  3. Zhang, Z., Wang, Y., Huang, T., Zhan, L.: A weighting algorithm based on the gravitational model for local stereo matching. Signal Image Video Process. 14, 315–323 (2020)

    Article  Google Scholar 

  4. Yang, Q.: Stereo matching using tree filtering. IEEE Trans. Pattern Anal. Mach. Intell. 37, 834–846 (2015)

    Article  Google Scholar 

  5. Zhang, Z., Zhan, L., Wang, Y.: Outlier removal based on Chauvenet’s criterion and dense disparity refinement using least square. J. Electron. Image 28(2), 1–9 (2019)

    Google Scholar 

  6. Park, H., Lee, K.: Look wide to match image patches with convolutional neural networks. IEEE Signal Process. Lett. 24, 1788–1793 (2017)

    Article  Google Scholar 

  7. Zbontar, J., LeCun, Y.: Stereo matching by training a convolutional neural network to compare image patches. J. Mach. Learn. Res. 17, 1–32 (2016)

    MATH  Google Scholar 

  8. Ren, C., Sheng, C.: Pyramid stereo matching network. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pp. 2339–2348 (2018)

  9. Kong, D., Tao, H.: A method for learning matching errors for stereo computation. In: Proceedings of Conference the British Machine Vision, pp. 1–11 (2004)

  10. Merlin, C., Rejimol, R.: Novel stereo matching technique for radiometric changes using normalized cross correlation. In: Proceedings of Conference Data Science and Engineering, pp. 94–97 (2012)

  11. Egnal, G., Wildes, R.: Detecting binocular half-occlusions: empirical comparisons of five approaches. IEEE Trans. Pattern Anal. Mach. Intell. 24, 1127–1132 (2002)

    Article  Google Scholar 

  12. Bernhardt, S., Boutet, J., Duboeuf, F., et. al.: Bimodality ultrasound/optical imaging: registration and 2D/3D visualization. In: Proceedings of IEEE Conference Ultrasonics Symposium, pp. 2127–2130 (2010)

  13. Espinosa, N., Andreas, L., Wolfgang, C.: Towards Fast 3D Reconstruction of Urban Areas from Aerial Nadir Images for a Near Real-Time Remote Sensing System. In: Proceedings of IEEE Conference Geoscience and Remote Sensing Symposium, pp. 6472–6475 (2018)

  14. Jie, Z., Wang, P.,Li, Y.: Left-Right Comparative Recurrent Model for Stereo Matching. In: Proceedings of IEEE Conference Computer Vision and Pattern Recognition, pp. 3838–3846 (2018)

  15. Zhang, K., Li, J.: Binary stereo matching. In: Proceedings of IEEE Conference Pattern Recognition, pp. 356–359 (2012)

  16. Bricola, C., Bilodeau, M., Beucher, S.: A top-down approach to the estimation of depth maps driven by morphological segmentations. In: Proceedings of International Conference Symposium on Mathematical Morphology and its Applications to Signal and Image Processing, pp. 122–133 (2015)

  17. Lu, Z., Wang, J., Li, Z., Chen, S., Wu, F.: A resource-efficient pipelined architecture for real-time semi-global stereo matching. IEEE Trans. Circuits Syst. Video Technol. 99, 1–10 (2021)

    Google Scholar 

  18. Tian, X., Sun, R.: Stereo matching based on multi-direction polynomial model. Signal Process. Image Commun. 44, 44–56 (2016)

  19. Sinha, S., Scharstein, D., Szeliski, D.: Efficient high resolution stereo matching using local plane sweeps. In: Proceedings of International Conference Computer Vision and Pattern Recognition, pp. 1582–1589 (2014)

  20. Tian, X., Sun, R.: Stereo matching using cost volume watershed and region merging. Signal Process. Image Commun. 29(10), 1232–1244 (2014)

    Article  Google Scholar 

  21. Suykens, J., Lukas, L., Vandewalle, J.: Least squares support vector machine classifiers. Neural Process. Lett. 9, 293–300 (1999)

    Article  Google Scholar 

  22. Scharstein, D., Hirschmuller, H., Kitajima, K., et,al.: High resolution stereo datasets with subpixel accurate ground truth. In: Proceedings of International Conference German Conference on Pattern Recognition, pp. 1–12 (2014)

  23. Alessio, T., Matteo, P., Stefano, M., et al.: Unsupervised domain adaptation for depth prediction from images. IEEE Trans. Pattern Anal. Mach. Intell. 42(10), 2396–2409 (2020)

    Article  Google Scholar 

Download references

Funding

This study was funded by the science and technology project of science and technology department of Henan province (No. 212102210149) and the national natural science foundation of China under grants (No. 62006071). The key and promotion Projects of Henan Province (No. 212102210152); Natural science foundation of Xinjiang Uygur Autonomous Region (No. 2018D01C003). Henan University of technology high level talents research start up fund project (No. 2020BS023).

Author information

Authors and Affiliations

  1. School of Artificial Intelligence and Big Data, Henan University of Technology, Lianhua street, Zhengzhou, 450001, China

    Zhang zihao, Yu junwei & Mu yashuang

  2. School of Electronic Information Engineering, Yili Normal University, Yil, China

    Wang xuefeng

Authors
  1. Zhang zihao
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Wang xuefeng
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Yu junwei
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Mu yashuang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Zhang zihao.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

zihao, Z., xuefeng, W., junwei, Y. et al. Disparity refinement based on least square support vector machine for stereo matching. SIViP 16, 2141–2148 (2022). https://doi.org/10.1007/s11760-022-02176-5

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11760-022-02176-5

Keywords

Search

Navigation