Abstract.
Digital cameras normally sample one color at each pixel. Missing colors are obtained by spatial interpolation, decreasing resolution relative to images acquired with a greyscale sensor. The consequence for document imaging is higher text recognition error rates. This paper introduces the horizontal-vertical regression (HVR) method for document-optimized color reconstruction. HVR exploits a local two-color approximation, making spatial interpolation unnecessary. Comparison with the best alternative reconstruction methods indicates large reductions in error rates for text resulting from HVR, as well as improvements in intermediate color and binary images.
Similar content being viewed by others
References
Sony Corporation (2003) Realization of natural color reproduction in digital still cameras, closer to the natural sight perception of the human eye. Sony Technology Development Article No. 03-029E, July 2003
ScanSoft Inc. Omnipage Pro 11. http://www.scansoft.com/omnipage/
Adams JE Jr (1995) Interactions between color plane interpolation and other image processing functions in electronic photography. In: Proc. SPIE: Cameras and Systems for Electronic Photography and Scientific Imaging, vol 2416. International Society for Optical Engineering, January 1995, pp 144-151
Lee H-C (1990) Review of image-blur models in a photographic system using the principles of optics. Opt Eng 29(5):405-421
Lopresti D, Tomkins A (1997) Block edit models for approximate string matching. Theor Comput Sci 181:159-179
Lyon R, Hubel P (2002) Eyeing the camera: into the next century. In: Proc. IST/TSID 10th conference on color imaging, pp 349-255
Morishita M (1981) Color imaging pick-up apparatus. US Patent 4,282,547
Newman W, Dance C, Taylor A, Taylor S, Taylor M, Aldhous A (1999) CamWorks: A video-based tool for efficient capture from paper source documents. In: Proc. IEEE international conference on multimedia computing and systems
Parulski K (1985) Color filters and processing alternatives for one-chip cameras. IEEE Trans Electron Devices 32(8):1381-1389
Pilu M, Pollard S (2002) A light-weight text image processing method for binarising camera images for OCR. In: Proc. British machine vision conference, pp 54-59
Ramanath R, Snyder WE (2003) Adaptive demosaicking. J Electron Imag 12(4)633-642
Ramanath R, Snyder WE, BilbroGL, Sander WA III (2002) Demosaicking methods for bayer color arrays. J Electron Imag 11(3):306-315
Savakis AE, Trussell HJ (1993) On the accuracy of PSF representation in image restoration. IEEE Trans Image Process 2:252-259
Seeger M, Dance C (2001) Binarising camera images for OCR. In: Proc. 6th international conference on document analysis and recognition, pp 54-59
Sharma G (2001) Show-through cancellation in scans of duplex printed documents. IEEE Trans Image Process 10(5):736-754
Wilson RG (1994) Modelling and calibration of automated zoom lenses. PhD thesis, Robotics Insitute, Carnegie Mellon University, Pittsburgh
Author information
Authors and Affiliations
Corresponding author
Additional information
Received: 11 June 2003, Accepted: 6 March 2004, Published online: 2 February 2005
C. Dance: Correspondence to
Rights and permissions
About this article
Cite this article
Dance, C., Fan, L. Color reconstruction in digital cameras: optimization for document images. IJDAR 7, 138–146 (2005). https://doi.org/10.1007/s10032-004-0130-7
Issue Date:
DOI: https://doi.org/10.1007/s10032-004-0130-7