COSCH Session at the Computational Color Imaging Workshop 2015

Alain Trémeau reports

A specialist session, Color in Digital Cultural Heritage, was held during the 5th Computational Color Imaging Workshop (CCIW'2015) at the Jean Monnet University in St Etienne, France, on 26 March 2015. The session was organised in cooperation with Colour and Space in Cultural Heritage (COSCH), the COST Action TD 1201. What follows is a brief report on different papers presented during this expert session. They are presented in context of other papers on the state of the art, published recently.

The processes of colour/multispectral digitisation of 3D objects of cultural heritage have recently received much attention, mainly due to the quality and resolution of digital objects that are still improving [1, 2, 3].

One way to improve colour accuracy and to perceive invisible information (i.e. outside the visible range) is to use a multispectral system and to fuse, for example, colour and infra-red. Thus, Jay Arre Toque et al. propose to use a high-resolution multispectral scanning in order to analyse the discolouration of traditional Japanese pigments [4]. Other authors propose to use a hyperspectral camera for examination of paintings, e.g. [5].

Another tendency is to improve the performance of colour-difference formulas [6] or to evaluate with these formulas if a colour digitization system is accurate enough. Thus, Tatiana Vitorino et al. propose to use the ColorChecker chart to assess the usefulness and comparability of data acquired with two hyper spectral systems [7]. Independently of this paper, several member laboratories of the COSCH Working Group 1 (WG1) are currently performing colour/spectral measurements and comparisons to assess the accuracy of several instruments in terms of spectral estimation and colour accuracy from the ColorChecker chart, and also from other colour samples (COSCH Round Robin Test). Other  authors propose to use a specular colour chart, rather than the ColorChecker chart, to calibrate the colour digitisation of highly specular material systems [8]. Another challenge is related to the colour rendering in reproduction and visualisation. Indeed, the problems investigated in the digital cultural heritage community define an area of research on the frontier between vision science and image/display technology which has to be addressed from several different perspectives, involving disciplines such as physics, psychophysics, artificial intelligence and fine arts. According to J.J. McCann, the colour appearance of a scene is not only correlated to the surfaces reflectance (which depends on the spectral distribution and the dynamic range of the illumination) and to the scene spatial content (e.g., a flat representation or a 3D scene), but also to the sensitivities of the human L, M, S cones [9, 10].

The process of 3D object visualisation in cultural heritage has also received much attention due to the development of colour rendering and colour correction algorithms. Some authors studied these issues from the observer's perspective (i.e. visual observation). Thus, Sergio Nascimento et al. investigated which colour compositions observers prefer when they look at some paintings [11]. Other authors proposed to address these issues using photometry/spectrophotometry models. For example, in his invited presentation, Lindsay MacDonald addressed problems related to the modelling of the diffused and specular reflectance of 3D objects, and to the 3D surface reconstruction from photometric stereo [12]. Another approach is to address these questions using computer vision models. For example, Zoltan Kato in his invited presentation discussed problems related to the pose estimation without the use of any special calibration pattern or explicit point correspondence [13]. This paper addressed one of the most challenging issues in digital cultural heritage, which is the fusion of heterogeneous data, such as 2D RGB/spectral imagery with other 3D range sensing modalities (e.g., Lidar). Kato proposes to select manually or automatically a few planar areas in 2D and 3D images and match them geometrically. When some images are textureless (e.g. 3D images) this method performs well, while methods based on detection of interest points are irrelevant. Citlalli Gamez Serna et al., on the other hand, propose a semi-automatic 2D-3D registration framework to produce accurate realistic results from a set of 2D uncalibrated images and a sparse 3D point cloud representation of an object digitised with laser scanning [14].

References

  1. Karaszewski, M., Lech, K., Bunsch, E. and Sitnik, R., 2014. In the pursuit of perfect 3D digitization of surfaces of paintings: geometry and color optimization. In: Proceedings of the 5th International Conference EuroMed'2014, Limassol, Cyprus, 3–8 November, pp. 25–34.
  2. Boochs, F., Bentkowska-Kafel, A., Degrigny, Ch., Karaszewski, M., Karmacharya, A., Kato, Z., Picollo, M., Sitnik, R. and Trémeau, A., 2014. Colour and Space in Cultural Heritage. Key questions in 3D optical documentation of material culture for conservation, study and preservation. In: Proceedings of the 5th International Conference EuroMed 2014, Best paper award, Limassol, Cyprus, 3–8 November, pp. 11–14.
  3. Boochs, F., Trémeau, A., Murphy, O., Gerke, M., Lerma, J.L., Karmacharya, A. and Karaszewski, M., 2014. Towards a knowledge model bridging technologies and applications in cultural heritage documentation. In: Proceedings of the ISPRS Technical Commission V Symposium, Riva del Garda, Italy, 23–25 June, pp. 81–88.
  4. Toque, J.A., Zhan, P., Wang, P. and Ide-Ektessabi, A., 2015. High-resolution multispectral scanning for mesoscopic investigation of discoloration of traditional Japanese pigments. In: Proceedings of the 5th International Workshop CCIW'2015, Jean Monnet University, St Etienne, France, 24–26 March. Springer International Publishing.
  5. Chen, A., 2013. Colour Visualisation of Hyperspectral Images in Art Restoration. CIMET Master Thesis, Gjøvik University College, Norway.
  6. Melgosa, M., Trémeau, A. and Cui, G., 2013. Colour Difference Evaluation. In: Advanced Color Image Processing and Analysis, C. Fernandez-Maloigne eds., pp. 59–79.
  7. Vitorino, T., Casini, A., Cucci, C., Gebejes, A., Hiltunen, J., Hauta-Kasari, M., Picollo, M. and Stefani, L. 2015. Accuracy in colour reproduction: using a ColorChecker chart to assess the usefulness and comparability of data acquired with two hyper-spectral systems. In: Proceedings of the 5th International Workshop CCIW'2015. Jean Monnet University, St Etienne, France, 24–26 March. Springer International Publishing.
  8. Martínez-García, J., Hébert, M. and Trémeau, A., 2015. Color calibration of an RGB digital camera for the microscopic observation of highly specular materials. In: Proceedings of the SPIE conference on Measuring, Modeling, and Reproducing Material Appearance, 9–10 February. San Francisco, CA.
  9. McCann, J. J. and Rizzi, A. 2011. The Art and Science of HDR Imaging, Wiley-IS&T Series in Imaging Science and Technology, ISBN: 978-0-470-66622-7, 416 pages.
  10. McCann, J. J., 2013. Spatial imaging in color and HDR: Prometheus unchained, Proceedings of SPIE, Human Vision and Electronic Imaging xviii, B. E. Rogowitz/T.N. Pappas/H. de Ridder, eds., Vol. 8651, 865107.
  11. Nascimento, S. M.C., Linhares, J.M.M., João, C.A.R., Amano, K., Montagner, C., Melo, M.J. and Vilarigues, M., 2015. Estimating the colors of paintings. In: Proceedings of the 5th International Workshop CCIW'2015. Jean Monnet University, St Etienne, France, 24–26 March. Springer International Publishing.
  12. MacDonald, L., 2015. Representation of Cultural Objects by Image Sets with Directional Illumination. In: Proceedings of the 5th InternationalWorkshop CCIW'2015. Jean Monnet University, St Etienne, France, 24–26 March. Springer International Publishing.
  13. Kato, Z. and Tamas, L., 2015. Relative Pose Estimation and Fusion of 2D Spectral and 3D Lidar Images. In: Proceedings of the 5th International Workshop CCIW'2015. Jean Monnet University, St Etienne, France, 24–26 March. Springer International Publishing.
  14. Serna, C.G., Pillay, R. and Trémeau, A., 2015. Data fusion of objects using techniques such as Laser Scanning, Structured Light and Photogrammetry for Cultural Heritage Applications. In: Proceedings of the 5th International Workshop CCIW'2015. Jean Monnet University, St Etienne, France, 24–26 March. Springer International Publishing.

 

Average (0 Votes)
No comments yet. Be the first.

Information Information

COSCH final book

General

FORTHCOMING: COSCH final book 

Digital Techniques for Documenting and Preserving Cultural Heritage

"The essays in this collection are transformative, moving beyond basic collaboration and skilfully contextualizing both scientic knowledge in the humanities and humanities knowledge in the sciences. Doing so not only heightens the quality of the research, but heightens understanding, redrawing traditional lines between disciplines and redening what it means to truly collaborate and to be a scholar in the digital age."-Bill Endres, University of Oklahoma 
In this unique collection the authors present a wide range of interdisciplinary methods to study, document, and conserve material cultural heritage. The methods used serve as exemplars of best practice with a wide variety of cultural heritage objects having been recorded, examined, and visualised. The objects range in date, scale, materials, and state of preservation and so pose dierent research questions and challenges for digitization, conservation, and ontological representation of knowledge. Heritage science and specialist digital technologies are presented in a way approachable by non-scientists, while a separate technical section provides details of methods and techniques, alongside examples of notable applications of spatial and spectral documentation of material cultural heritage, with selected literature and identication of future research. 
This book is an outcome of interdisciplinary research and debates conducted by the participants of the COST Action TD1201, Colour and Space in Cultural Heritage, 2012–16, and is an Open Access publication available under a CC BY-NC-ND licence.