Document Compression Using H.264/AVC

It has been verified that H.264/AVC, the newest video compression standard, can also be used to encode still images. In many cases, it outperforms state-of-art coders such as JPEG2000. For compound documents, the gains over JPEG2000 are even more expressive. In this scenario, the contributions of the present paper are distributed over four document encoding methods that use the H.264/AVC as a basic functional element, namely: method 1, Advanced Video Coding Compound, which, based on a macroblock content analysis, adaptively encodes text and image regions; method 2, MRC Compression of Electronically Generated Documents using H.264/AVC-I and JBIG2, which combines MRC (Mixed Raster Content) with H.264/AVC and JBIG2, and proposes a new data-filling technique based on the H.264/AVC intra prediction; method 3, MRC Compression of Scanned Documents using H.264/AVC-I and JBIG2, which offers pre/post-processing techniques as extensions of the MRC imaging model; and method 4, Compression of Scanned Books using H.264/AVC, which explores pattern recurrence to encode pages of scanned books. Many experiments were carried out in order to verify the efficiency of the proposed methods. Results showed objective and/or subjective gains over known approaches. Keywords-Image processing; compound document compression; Mixed Raster Document; H.264/AVC; scanned book compression.