Selected Papers on Image Processing and Image Analysis

date:July 7, 2007

Low-level image processing

  • Implementation efficiency of binary morphology (in pdf or gzipped ps), D. S. Bloomberg, Mathematical Morphology and its applications to image and signal processing: Proceedings of the Sixth International Symposium on Mathematical Morphology, pp. xx, CSIRO Pub, April 2002. An informal introduction and overview of this paper exists (in pdf or gzipped ps).

  • Pattern matching using the blur hit-miss transform (in pdf or gzipped ps), D. S. Bloomberg and L. Vincent, Journal of Electronic Imaging, Vol 9(2), pp. 140–150, April 2000.

  • Connectivity-preserving morphological image transformations (in pdf or gzipped ps), D. S. Bloomberg, SPIE Conf. 1606, Visual Communications and Image Processing \‘91, pp. 320–334, Boston, MA, November 11–13, 1991.

  • Generalized hit-miss operations (in pdf or gzipped ps), D. S. Bloomberg and P. Maragos, SPIE Conf. 1350, Image Algebra and Morphological Image Processing, pp. 116–128, San Diego, CA, July 9–13, 1990. This paper is a slightly revised version, with an added figure.

  • Generalized hit-miss operators with applications to document image analysis (in pdf or gzipped ps), D. S. Bloomberg and P. Maragos. This is an unpublished version of the SPIE paper above, with a significantly expanded section on blur and rank order versions of the hit-miss operator.

Basic methods of image segmentation

  • Textured reductions for document image analysis (in pdf or gzipped ps), D. S. Bloomberg, SPIE Conf. 2660, Doc. Rec. III, pp. 160–174, San Jose, CA, Jan. 30–31, 1996.

  • Image analysis using threshold reduction (in pdf or gzipped ps), D. S. Bloomberg, SPIE Conf. 1568, Image Algebra and Morphological Image Processing II, pp. 38–51, San Diego, CA, July 23–24, 1991.

  • Multiresolution morphological approach to document image analysis (in pdf or gzipped ps), D. S. Bloomberg, International Conference on Document Analysis and Recognition, pp. 963–971, Saint-Malo, France, Sept. 30–Oct 2, 1991.

Document Image Decoding

  • An iterative algorithm for optimal message recognition in linguistically constrained document image decoding (in pdf), K. Popat, D. S. Bloomberg and D. Greene, Proceedings of the 4th IAPR Workshop on Document Analysis Systems, Springer, 2002.

  • Document image decoding using iterated complete path search with subsampled heuristic scoring (in pdf or gzipped ps), D. S. Bloomberg, T. P. Minka and K. Popat, Proceedings of the IAPR 2001 International Conference Document Analysis and Recognition (ICDAR 2001), September 2001.

  • Document image decoding using iterated complete path heuristic (in pdf or gzipped ps), T. Minka, D. S. Bloomberg and K. Popat, SPIE Conf. 4307, Document Recognition and Retrieval VIII, pp. 251–258, San Jose, CA, Jan. 24–25, 2001.

  • Adding linguistic constraints to document image decoding: comparing the iterated complete path and stack algorithms, K. Popat, D. Greene, J. Romberg and D. S. Bloomberg, SPIE Conf. 4307, Document Recognition and Retrieval VIII, pp. 259–271, San Jose, CA, Jan. 24–25, 2001.

Image Compression

  • Two-stage lossy/lossless compression of grayscale document images, (in pdf), K. Popat and D. S. Bloomberg, Mathematical Morphology and its applications to image and signal processing: Proceedings of the Fifth International Symposium on Mathematical Morphology, pp. 361–370, Kluwer Acad. Pub, June 2000.

  • Google Books: Making the public domain universally accessible (in pdf), A. Langley and D. S. Bloomberg, SPIE Conf. 6500, Doc. Rec. and Retrieval XIV, paper 6500-16, San Jose, CA, Jan 30–Feb 1, 2007.

Simple Document Image Applications

  • Determining the resolution of scanned document images (in pdf or gzipped ps), D. S. Bloomberg, SPIE Conf. 3651, Doc. Rec. VI, pp. 10–21, San Jose, CA, Jan 27–28, 1999.

  • Identifying document image skew and orientation (in pdf or gzipped ps), D. S. Bloomberg, G. E. Kopec and L. Dasari, SPIE Conf. 2422, Doc. Rec. II, pp. 302–316, San Jose, CA, Feb. 6–7, 1995.

Complex Document Applications using Segmentation

  • Summarization of imaged documents without OCR, F. R. Chen and D. S. Bloomberg, CVIU, Vol 70, No 3, pp. 307–320, 1998.

  • Reading digital data embedded in iconic text, D. S. Bloomberg, SPIE Conf. 3305, Doc. Rec. V, pp. 194–207, San Jose, CA, Jan. 28–29, 1998.

  • Extraction of indicative summary sentences from imaged documents, F. R. Chen and D. S. Bloomberg, ICDAR \‘97, pp. 227–232, Ulm, Germany, Aug. 18–20, 1997.

  • Image-based document summarization, F. R. Chen and D. S. Bloomberg, Symp. Doc. Image Under. Tech, Annapolis, MD, Apr. 30–May 2, 1997.

  • Embedding digital data on paper in iconic text, D. S. Bloomberg, SPIE Conf. 3027, Doc. Rec. IV, pp. 67–80, San Jose, CA, Feb. 12–13, 1997.

  • Document image summarization without OCR, D. S. Bloomberg and F. R. Chen, Int. Conf. Image Proc. 1996, Vol. II, pp. 229–232, Lausanne, Switzerland, Sept. 16–19, 1996.

  • Extraction of text-related features for condensing image documents, D. S. Bloomberg and F. R. Chen, SPIE Conf. 2660, Doc. Rec. III, pp. 72–88, San Jose, CA, Jan. 30–31, 1996.

Papers by Luc Vincent

Luc Vincent has made available a large number of his outstanding published papers, all in DjVu format. (DjVu is a “mixed raster” format, with separate text/line-graphics and image layers, both of which are compressed using lossy methods. The text layer is compressed by “tokenization”: the connected components are identified and those that are nearly indistinguishable visually are grouped together by an unsupervised classifier. Only one of these tokens from each class is used to reproduce the image; in the file it is compressed by an arithmetic coding scheme. Wavelet compression is used on the image parts. The result is the finest document image compression method — from a rate-distortion perspective — available.)

Comprehensive bibliography

There is a comprehensive bibliography of “computer vision”, image processing, and related topics.