Advertisement

Multimedia Tools and Applications

, Volume 77, Issue 23, pp 30615–30631 | Cite as

Efficient road specular reflection removal based on gradient properties

  • Yao Wang
  • Fangfa Fu
  • Fengchang Lai
  • Weizhe Xu
  • Jinjin Shi
  • Jinxiang Wang
Article
  • 97 Downloads

Abstract

Highlights caused by changes in sunlight throughout any given day cause failure in stereo matching, object recognition, and road segmentation. This is a serious challenge in advanced driver assistance systems (ADAS), because local high brightness and color discontinuities generally result in noticeable blurring of the road surface or object. This paper presents a novel strategy for removing specular reflection from highlight images by gradients distribution to optimize the diffuse image. The dark channel is introduced as a prior to initially estimate and locate the highlight. The threshold filter is then adopted to divide the high-intensity highlight and the weak highlight - the weak highlight affect neither the stereo matching nor road segmentation process. Finally, gradient properties (varying smoothness of specular and diffuse reflections) are presented to optimize the layer separation. Experimental results in speed and accuracy of road segmentation show that proposed method outperforms other techniques for separating highlights from road surfaces.

Keywords

Road segmentation ADAS Highlight removal Threshold filter Layer separation 

Notes

Acknowledgments

This work was supported by a grant from the National Natural Science Foundation of China (NSFC, No. 61504032)

References

  1. 1.
    Chung H-S (2008) Jiayajia efficient photometric stereo on glossy surfaces with wide specular lobes. In: Computer vision and pattern recognition, pp 1–8Google Scholar
  2. 2.
    Everingham M, Van Gool L, Williams CKI, Winn J, Zisserman A The PASCAL visual object classes challenge 2012 (VOC2012) Results. http://www.pascalnetwork.org/challenges/VOC/voc2012/workshop/index.html
  3. 3.
    Geiger A, Lenz P, Urtasun R (2012) Are we ready for autonomous driving? The kitti vision benchmark suite[C]. In: Computer vision and pattern recognition (CVPR), pp 3354–3361Google Scholar
  4. 4.
    Kim H, Jin H, Hadap S et al (2013) Specular reflection separation using dark channel prior[C]. In: Computer Vision and Pattern Recognition (CVPR), pp 1460–1467Google Scholar
  5. 5.
    Klinker GJ, Shafer SA, Kanade T (1988) The measurement of highlights in color images. Int J Comput Vis 2(1):7–32CrossRefGoogle Scholar
  6. 6.
    Klinker GJ, Shafer SA, Kanade T (1988) Image segmentation and reflection analysis through color[C]. In: Applications of artificial intelligence, pp 229–244Google Scholar
  7. 7.
    Lee SW, Bajcsy R (1992) Detection of specularity using color and multiple views. In: Proceedings of the 2nd Eur. Conf. Comput. Vis., pp 99–114CrossRefGoogle Scholar
  8. 8.
    Levin A, Weiss Y (2007) User assisted separation of reflections from a single image using a sparsity prior. IEEE Trans Pattern Anal Mach Intell 29(9):1647–1654CrossRefGoogle Scholar
  9. 9.
    Li Y, Brown MS (2014) Single image layer separation using relative smoothness[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 2752–2759Google Scholar
  10. 10.
    Lin S, Li Y, Kang SB et al (2002) Diffuse-specular separation and depth recovery from image sequences[C]. In: European conference on computer vision, pp 210–224CrossRefGoogle Scholar
  11. 11.
    Long J, Shelhamer E, Darrell T (2015) Fully convolutional networks for semantic segmentation. CVPR 3:4Google Scholar
  12. 12.
    Mallick SP, Zickler T, Kriegman DJ, Belhumeur PN (2005) Beyond Lambert: reconstructing specular surfaces using color. Comput Vis Pattern Recognit 2:619–626Google Scholar
  13. 13.
    Nayar SK, Fang XS, Boult T (1997) Separation of reflection components using color and polarization. Int J Comput Vis 21(3):163–186CrossRefGoogle Scholar
  14. 14.
    Oliveira GL, Burgard W, Brox T (2016) Efficient deep models for monocular road segmentation[C]. In: Intelligent robots and systems (IROS), pp 4885–4891Google Scholar
  15. 15.
    Park JS, Tou JT (1990) Highlight separation and surface orientation for 3-d specular objects. Pattern Recogn 1:331–335Google Scholar
  16. 16.
    Scharstein D, Szeliski R (2002) A taxonomy and evaluation of dense two-frame stereo correspondence algorithms[J]. Int J Comput Vis 47(1-3):7–42CrossRefGoogle Scholar
  17. 17.
    Shi J, Fu F, Wang Y et al (2016) Stereo matching with improved radiometric invariant matching cost and disparity refinement[C]. In: International conference on intelligent computing, pp 61–73Google Scholar
  18. 18.
    Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. CoRR, arXiv:1409.1556
  19. 19.
    Tan RT, Ikeuchi K (2005) Separating reflection components of textured surfaces using a single image[J]. Pattern Anal Mach Intell 27(2):178–193CrossRefGoogle Scholar
  20. 20.
    Tan RT, Ikeuchi K (2005) Reflection components decomposition of textured surfaces using linear basis functions[C]. In: Computer vision and pattern recognition, pp 125–131Google Scholar
  21. 21.
    Tan P, Quan L, Lin S (2006) Separation of highlight reflection on textured surfaces. Comput Vis Pattern Recognit 2:1855–1860Google Scholar
  22. 22.
    Teichmann M, Weber M, Zoellner M et al (2016) MultiNet: real-time joint semantic reasoning for autonomous driving[J]. arXiv:1612.07695
  23. 23.
    Tomasi C, Manduchi R (1998) Bilateral filtering for gray and color images. In: Computer vision, pp 839–846Google Scholar
  24. 24.
    Wang Y, Fu F, Shi J et al (2016) Efficient specular reflection separation based on dark channel prior on road Surface[C]. In: International conference on intelligent computing. Springer, Cham, pp 426–435CrossRefGoogle Scholar
  25. 25.
    Wolff LB, Boult TE (1991) Constraining object features using a polarization reflectance model[J]. IEEE Trans Pattern Anal Mach Intell 13(7):635–657CrossRefGoogle Scholar
  26. 26.
    Yan C, Zhang Y, Xu J et al (2014) Efficient parallel framework for HEVC motion estimation on many-core processors[J]. IEEE Trans Circuits Syst Video Technol 24(12):2077–2089CrossRefGoogle Scholar
  27. 27.
    Yan C, Zhang Y, Xu J et al (2014) A highly parallel framework for HEVC coding unit partitioning tree decision on many-core processors[J]. IEEE Signal Process Lett 21(5):573–576CrossRefGoogle Scholar
  28. 28.
    Yan C, Xie H, Liu S et al (2018) Effective Uyghur language text detection in complex background images for traffic prompt identification[J]. IEEE Trans Intell Transp Syst 19(1):220–229CrossRefGoogle Scholar
  29. 29.
    Yan C, Xie H, Yang D et al (2018) Supervised hash coding with deep neural network for environment perception of intelligent vehicles[J]. IEEE Trans Intell Transp Syst 19(1):284–295CrossRefGoogle Scholar
  30. 30.
    Yang Q, Tan K-H, Ahuja N (2009) Real-time o(1) bilateral filtering. In: Computer vision and pattern recognition, pp 557–564Google Scholar
  31. 31.
    Yang Q, Wang S, Ahuja N (2010) Real-time specular highlight removal using bilateral filtering[C]. In: European conference on computer vision. Springer. Berlin, pp 87–100CrossRefGoogle Scholar
  32. 32.
    Yang Q, Tang J, Ahuja N (2015) Efficient and robust specular highlight removal [J]. IEEE Trans Pattern Anal Mach Intell 37(6):1304–1311CrossRefGoogle Scholar
  33. 33.
    Zeiler MD, Fergus R (2014) Visualizing and understanding convolutional networks. In: European conference on computer vision, pp 818–833Google Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  • Yao Wang
    • 1
  • Fangfa Fu
    • 1
  • Fengchang Lai
    • 1
  • Weizhe Xu
    • 1
  • Jinjin Shi
    • 2
  • Jinxiang Wang
    • 1
  1. 1.Microelectronics CenterHarbin Institute of TechnologyHarbinChina
  2. 2.College of Mechanical and Power EngineeringChina Three Gorges UniversityYichangChina

Personalised recommendations