KIPS Transactions on Software and Data Engineering (정보처리학회논문지:소프트웨어 및 데이터공학)

Korea Information Processing Society (한국정보처리학회)
권호 표지
DOI QR코드

DOI QR Code

Estimation of Illuminant Chromaticity by Equivalent Distance Reference Illumination Map and Color Correlation

균등거리 기준 조명 맵과 색 상관성을 이용한 조명 색도 추정

  • Received : 2022.08.03
  • Accepted : 2022.12.28
  • Published : 2023.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, a method for estimating the illuminant chromaticity of a scene for an input image is proposed. The illuminant chromaticity is estimated using the illuminant reference region. The conventional method uses a certain number of reference lighting information. By comparing the chromaticity distribution of pixels from the input image with the chromaticity set prepared in advance for the reference illuminant, the reference illuminant with the largest overlapping area is regarded as the scene illuminant for the corresponding input image. In the process of calculating the overlapping area, the weights for each reference light were applied in the form of a Gaussian distribution, but a clear standard for the variance value could not be presented. The proposed method extracts an independent reference chromaticity region from a given reference illuminant, calculates the characteristic values in the r-g chromaticity plane of the RGB color coordinate system for all pixels of the input image, and then calculates the independent chromaticity region and features from the input image. The similarity is evaluated and the illuminant with the highest similarity was estimated as the illuminant chromaticity component of the image. The performance of the proposed method was evaluated using the database image and showed an average of about 60% improvement compared to the conventional basic method and showed an improvement performance of around 53% compared to the conventional Gaussian weight of 0.1.

본 논문에서는 입력 영상에 대한 촬영 장면의 조명 색도를 추정하는 방법을 제안한다. 조명 기준영역을 이용하여 입력영상의 촬영 장면에 가장 근접한 조명 색도를 추정한다. 기존의 방법은 일정한 수의 기준조명 정보를 이용한다. 입력 영상으로부터 화소의 색도분포 정보와 기준 조명에 대한 미리 준비된 색도 집합을 대조하여 겹치는 면적이 가장 큰 기준 조명을 해당 입력 영상에 대한 장면 조명으로 간주한다. 겹치는 면적을 계산하는 과정에서 각 기준 조명에 대한 가중치를 가우시안 분포 형태로 적용하였으나, 분산 값에 대하여 명확한 기준을 제시하지 못하였다. 제안한 방법은 주어진 기준조명으로부터 독립적인 기준색도 영역을 추출하고, 입력영상의 모든 화소에 대하여 RGB 칼라좌표계의 r-g 색도 평면에서의 특징치를 계산한 다음, 독립적인 색도영역과 입력영상으로부터의 특징치를 이용하여 유사도를 평가한다. 유사도가 가장 높게 나타나는 조명을 해당 영상의 조명 색도 성분으로 추정하였다. 데이터베이스의 영상과 기준조명 색도를 이용한 성능평가에서 제안한 방법은 기존의 기본 방법에 비하여 평균 60% 정도의 개선을 보였고, 기존의 가우시안 분산 값이 0.1인 경우에 비하여 53% 내외의 개선 성능을 보였다.

Keywords

Acknowledgement

이 논문은 2022학년도 영산대학교 교내연구비의 지원에 의하여 이루어진 것임.

References

  1. E. H. Land, "Color vision and the natural image," Proceedings of the National Academy of Sciences, Vol.45, pp.115-129, 1959; Vol.45, pp.636-644, 1959; Vol.80, pp.5163-5169, 1983. https://doi.org/10.1073/pnas.80.16.5163
  2. G. J. Klinker, S. A. Shafer, and T. Kanade, "The Measurement of Highlights in Color Images," International Journal of Computer Vision, Vol.2, No.1, pp.7-32, 1988. https://doi.org/10.1007/BF00836279
  3. H. C. Lee, "Method for computing the scene-illuminant chromaticity from specular highlights," Journal of the Optical Society of America, Vol.3, No.10, pp.1694-1699, 1986. https://doi.org/10.1364/JOSAA.3.001694
  4. B. Li, D. Xu, W. Xiong, and S. Feng, "Color constancy using achromatic surface," Color Research and Application, Vol.35, Iss.4, pp.304-312, 2010. https://doi.org/10.1002/col.20574
  5. F. Chang, S. Pei, and W. Chao, "Color constancy by chromaticity neutralization," Journal of Optical Society of America A, Vol.29, No.10, pp.2217-2225, 2012. https://doi.org/10.1364/JOSAA.29.002217
  6. H. Joze, M. Drew, G. Finlayson, and P. Troncoso, "The role of bright pixels in illumination estimation," Final Program and Proceedings - IS and T/SID Color Imaging Conference, pp.41-46, 2012.
  7. L. T. Maloney and B. A. Wandell, "A computational model for color constancy," Journal of the Optical Society of America, Vol.3, pp.29-33, 1986. https://doi.org/10.1364/JOSAA.3.000029
  8. D. A. Forsyth, "A novel algorithm for color constancy," International Journal of Computer Vision, Vol.5, No.1, pp.5-36, 1990. https://doi.org/10.1007/BF00056770
  9. G. Finlayson, S. Hordley, and P. Hubel, "Color by correlation: A simple, unifying framework for color constancy," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.23, No.11, pp.1209-1221, 2001. https://doi.org/10.1109/34.969113
  10. K. Barnard, C. Vlad, and B. Funt, "A comparison of computational color constancy algorithms. I: Methodology and experiments with synthesized data," IEEE Transactions on Image Processing, Vol.11, No.9, pp.972-984, 2002. https://doi.org/10.1109/TIP.2002.802531
  11. K. Barnard, L. Martin, A. Coath, and B. Funt, "A comparison of computational color constancy algorithms? Part II: Experiments with image data," IEEE transactions on Image Processing, Vol.11, No.9, pp.985-996, 2002. https://doi.org/10.1109/TIP.2002.802529
  12. W. Xiong and B. Funt, "Estimating illumination chromaticity via support vector regression," Journal of Imaging Science and Technology, Vol.50, No.4, pp.341-348, 2006. https://doi.org/10.2352/J.ImagingSci.Technol.(2006)50:4(341)
  13. V. Agarwal, A. Gribok, and M. Abidi, "Machine learning approach to color constancy," Neural Networks, Vol.20, Iss.5, pp.559-563, 2007. https://doi.org/10.1016/j.neunet.2007.02.004
  14. S. Bianco, C. Cusano, and R. Schettini, "Color constancy using CNNs," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015.
  15. W. Shi, C. Loy, and X. Tang, "Deep specialized network for illuminant estimation," In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds) Computer Vision, European Conference on Computer Vision 2016, Vol.9908. pp.371-387, Springer, Cham, 2016.
  16. S. Oh and S. Kim, "Approaching the computational color constancy as a classification problem through deep learning," Pattern Recognition, Vol.61, pp.405-416, 2017. https://doi.org/10.1016/j.patcog.2016.08.013
  17. K. Koscevic, M. Subasic, and S. Loncaric, "Deep learningbased illumination estimation using light source classification," IEEE Access, Vol.8, pp.84239-84247, 2020 https://doi.org/10.1109/ACCESS.2020.2992121
  18. H. Choi, H. Kang, and B. Yun, "CNN-based illumination estimation with semantic information," Applied Sciences, Vol.10, No.14, Article-Number 4806, 2020.
  19. J. Qiu, H. Xu, and Z. Ye, "Color constancy by reweighting image feature maps," IEEE Transactions on Image Processing, Vol.29, pp.5711-5721, 2020. https://doi.org/10.1109/TIP.2020.2985296
  20. Y. Luo, X. Wang, Q. Wang, and Y. Chen, "Illuminant estimation using adaptive neuro-fuzzy inference system," Applied Sciences, Vol.11, No.21, Article-Number 9936, 2021.
  21. X. Zhang, S. Yue, R. Gong, and Q. Li, "Illumination estimation based on a weighted color distribution," Optik, Vol.185, pp.965-971, 2019. https://doi.org/10.1016/j.ijleo.2019.04.001
  22. P. V. Gehler, C. Rother, A. Blake, T. Sharp, and T. Minka, "Bayesian color constancy revisited," Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2008.
  23. E. Ershov et al., "The Cube++ Illumination Estimation Dataset," IEEE Access, pp.1-1, 2020.
  24. J. Kim and N. Kim, "Adversarial Learning-Based Image Correction Methodology for Deep Learning Analysis of Heterogeneous Images," KIPS Transactions on Software and Data EngineeringVol.10, No.11, pp.457-464, 2021.