Identifying the existence of bar codes in compressed images
CVGIP: Graphical Models and Image Processing
Computer vision: compress to comprehend
Pattern Recognition Letters
Document image compression and analysis
Document image compression and analysis
Spatial Sampling of Printed Patterns
IEEE Transactions on Pattern Analysis and Machine Intelligence
Managing Gigabytes: Compressing and Indexing Documents and Images
Managing Gigabytes: Compressing and Indexing Documents and Images
Lossless and lossy compression of text images by soft pattern matching
DCC '96 Proceedings of the Conference on Data Compression
An OCR based on character shape codes and lexical information
ICDAR '95 Proceedings of the Third International Conference on Document Analysis and Recognition (Volume 2) - Volume 2
An overview of the basic principles of the Q-Coder adaptive binary arithmetic coder
IBM Journal of Research and Development - Q-Coder adaptive binary arithmetic coder
Probability estimation for the Q-Coder
IBM Journal of Research and Development - Q-Coder adaptive binary arithmetic coder
Hi-index | 0.00 |
Document images belong to a unique class of images where theinformation is embedded in the language represented by a series ofsymbols on the page rather than in the visual objectsthemselves. Since these symbols tend to appear repeatedly, adomain-specific image coding strategy can be designed to facilitateenhanced compression and retrieval. In this paper we describe a codingmethodology that not only exploits component-level redundancy toreduce code length but also supports efficient data access. Theapproach identifies and organizes symbol patterns which appearrepeatedly. Similar components are represented by a single prototypestored in a library and the location of each component instance iscoded along with the residual between it and its prototype. Arepresentation is built which provides a natural information indexallowing access to individual components. Compression results arecompetitive and compressed-domain access is superior to competingmethods. Applications to network-related problems have beenconsidered, and show promising results.