Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Lossless Compression of Textual Images: A Study on Indic Script Documents
Ist Teil von
18th International Conference on Pattern Recognition (ICPR'06), 2006, Vol.3, p.806-809
Ort / Verlag
IEEE
Erscheinungsjahr
2006
Quelle
IEEE Xplore
Beschreibungen/Notizen
This paper presents a method for lossless compression of Indian language textual images. The study is an extension of the previously developed pattern matching and substitution (PM&S)-based method for lossy compression of similar images. Here an efficient method for residue coding is proposed and its performance is compared with CCITT Gr-IV and JBIG. A set of 20 text images for two most popular Indic scripts, namely Devanagari (Hindi) and Bengali, is used in the experiment. It is noted that the best results is achieved by PM&S-based approach followed by LZW-based residue coding. This combined scheme gives lossless compression ratio of about 37.9