基于相机响应值预测的光谱重建方法研究

doi:10.3964/j.issn.1000-0593(2025)07-1809-11

摘要
参考文献
相关文章 (15)

全文: PDF (14760 KB)
输出: BibTeX | EndNote (RIS)

摘要：光谱是颜色信息的指纹，也是表征物质物化属性的重要特征，对于颜色高保真复制和物质识别分析具有重要应用价值。基于光谱重建的光谱成像技术利用成像系统拍摄物体表面数字图像，并通过光谱重建计算得到物体的光谱图像，能够克服传统RGB图像对成像条件的依赖性，更加精细地表征和识别物体。相对于价格昂贵的高光谱相机而言，该项技术能够有效提升光谱图像的空间分辨率和获取效率，降低设备成本。尽管该技术在实验室已经取得较好的进展，但其面向光源多变、光照不均、以及成像参数易变的开放环境应用时，仍面临许多挑战。光谱重建技术对成像条件变化敏感，以上因素导致成像条件的多变性，使得单一成像条件下构建的光谱重建模型难以成像多变的开放环境。为此，本研究提出一种基于相机raw响应值预测的光谱重建方法。方法以数码相机适应模型为基础，通过预测训练样本集在任意成像条件下的raw响应值，对相机进行光谱特性化建模，然后对相同成像条件下的测试对象进行光谱重建，能够有效克服开放环境下光谱重建面临的成像条件多变难题。研究以Colorchecker SG140色卡为训练样本，以Colorchecker 24色卡为测试样本，分别在实验室和开放环境的不同成像条件下对方法进行了测试。实验结果显示，在实验室不同成像条件下，利用预测响应值对拍摄响应值的光谱重建的平均RMSE为4.02%，平均色差ΔE₀₀为5.3；在开放环境不同成像条件下，利用预测响应值对拍摄响应值的光谱重建的平均RMSE为3.2%，平均色差ΔE₀₀为4.5，验证了所提方法在实验室和开放环境下的有效性，为开放场景下的光谱重建提供了一种可行的参考方案。此外，研究发现所提方法对使用的光谱重建算法敏感，不同的光谱算法在光谱和色度方面的表现存在一定差异。

关键词：光谱重建；开放环境；数码相机；成像模型；raw响应值预测

Abstract：The surface spectral reflectance of the object is regarded as the fingerprint of its color. At the same time, it is also important to characterize the physical and chemical properties of substances. Spectral imaging technology based on spectral reconstruction can overcome the dependence of RGB images on imaging conditions. Meanwhile, it can effectively improve the spectral image's spatial resolution and acquisition efficiency and reduce equipment costs. Different from the principle of multispectral cameras, spectral imaging based on spectral reconstruction first captures the digital images of the object using a digital imaging system, and then the corresponding spectral images are reconstructed using spectral reconstruction methods. Current research has made great achievements in the laboratory; however, dealing with rapidly changing light sources, illumination, and imaging parameters in an open environment presents significant challenges for spectral reconstruction. This is because a spectral reconstruction model established under one set of imaging conditions is not suitable for use under different imaging conditions. To deal with the challenges, we explored the feasibility of spectral reconstruction based on camera raw response prediction in this study. In the proposed method, the camera raw response of the training dataset under specific imaging conditions is first predicted via the camera imaging model, then the spectral reconstruction algorithm is applied to spectrally characterize the digital camera based on the training dataset, at last, the spectral reflectance of the testing target is reconstructed from the captured image under the same imaging condition. The study tested the prediction results of raw response values for ColorChecker SG 140 color cards in a closed-light box environment and an outdoor open environment under five different combinations of exposure time and ISO. Spectral reconstruction tests were also conducted using the ColorChecker SG 140 color card for the ColorChecker 24 color card. In the closed lightbox and under different imaging conditions, the average RMSE(%) for reconstructing the true photographic response values using the predicted response values was 4.02, with an average CIEDE2000 color difference of 5.3. In the outdoor open environment and under different imaging conditions, the average RMSE(%) was 3.2, with an average CIEDE2000 color difference of 4.5. The experimental results show that the proposed method still has good reconstruction accuracy in outdoor environments, providing a feasible reference solution for spectral reconstruction in an open environment. In addition, we find that the proposed method is sensitive to the spectral reconstruction algorithms used, and different algorithms have different performances in spectral and chromaticity aspects.

Key words：Spectral reconstruction; Open environment; Digital camera; Imaging model; Raw response prediction

收稿日期: 2024-09-12 修订日期: 2025-02-06

中图分类号:

O432.3

基金资助: 国家自然科学基金项目（62305255），湖北省自然科学基金项目（2022CFB537）资助

通讯作者: 彭涛 E-mail: pt@wtu.edu.cn

作者简介: 梁金星，1989年生，武汉纺织大学计算机与人工智能学院副教授 e-mail: jxliang@wtu.edu.cn

引用本文:

梁金星，周文森，胡鑫，李壹帆，王鼎康，李宁，彭涛. 基于相机响应值预测的光谱重建方法研究[J]. 光谱学与光谱分析, 2025, 45(07): 1809-1819.
LIANG Jin-xing, ZHOU Wen-sen, HU Xin, LI Yi-fan, WANG Ding-kang, LI Ning, PENG Tao. Research on Spectral Reconstruction Based on Camera Response Prediction. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2025, 45(07): 1809-1819.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2025)07-1809-11 或 https://www.gpxygpfx.com/CN/Y2025/V45/I07/1809

[1] Liang J, Xin L, Zuo Z, et al. Frontiers in Neuroscience, 2022, 16: 1031546.
[2] Zhao Y, Po L M, Yan Q, et al. Hierarchical Regression Network for Spectral Reconstruction From RGB Images. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020. 1695.
[3] Cai Y, Lin J, Lin Z, et al. MST++: Multi-stage Spectral-wise Transformer for Efficient Spectral Reconstruction. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. 745.
[4] Lin Yitun, Graham D Finlayson. Sensors, 2021, 21(16): 5586.
[5] Shrestha R, Hardeberg J Y. Optics Express, 2014, 22(8): 9123.
[6] Khan H A, Thomas J B, Hardeberg J Y, et al. JOSA A, 2017, 34(7): 1085.
[7] Khan H A, Thomas J B, Hardeberg J Y, et al. Optics Express, 2019, 27(2): 1051.
[8] Lin Y T, Finlayson G D. Exposure Invariance in Spectral Reconstruction From RGB Images. Color and Imaging Conference, Society for Imaging Science and Technology, 2019. 284.
[9] Lin Y T, Finlayson G D. Physically Plausible Spectral Reconstruction From RGB Images. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020. 532.
[10] Ibrahim A, Tominaga S, Horiuchi T. Optical Review 2011, 18(2): 231.
[11] LIANG Jin-xing, XIN Lei, CHENG Jing-yao, et al(梁金星, 辛磊, 程靖尧, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2023, 43(11): 3330.
[12] Ramanath R, Snyder W E, Yoo Y, et al. IEEE Signal Processing Magazine, 2005, 22(1): 34.
[13] Farrell J E, Catrysse P B, Wandell B A. Applied Optics, 2012, 51(4): A80.
[14] Qiu J, Xu H. Applied Optics, 2016, 55(25): 6989.
[15] Jiang J, LiuD, Gu J, et al. What is the Space of Spectral Sensitivity Functions for Digital Color Cameras? 2013 IEEE Workshop on Applications of Computer Vision (WACV), 2013. 168.
[16] Tominaga S, Nishi S, Ohtera R. Sensors, 2021, 21(15): 4985.
[17] Liang J, Wan X. Optics Express, 2017, 25(23): 28273.
[18] Connah D R, Hardeberg J Y. Spectral Recovery Using Polynomial Models. Color Imaging X: Processing, Hardcopy, and Applications. SPIE, 2005, 5667: 65.
[19] Cao B, Liao N, Cheng H. Color Research and Application, 2017, 42(3): 327.
[20] Xu P, Xu H, Diao C, et al. Applied Optics, 2017, 56(30): 8461.
[21] Luo M R, Cui G, Rigg B. Color Research and Application, 2001, 26(5): 340.