Scanned Compound Image Compression using
Block based Pattern Matching
Journal:
GRENZE International Journal of Engineering and Technology
Authors:
Shubhada S.Thakare, Sujata N.Kale, Swarup S. Mathurkar
Volume:
4
Issue:
1
Grenze ID:
01.GIJET.4.1.507_1
Pages:
7-13
Abstract
Scanned document compression is needed to store the conventional paper formats
in to digital formats. It can be implemented by digitizing scanned document and then
optimized memory through compression process. Scanned document may contain different
parts such as text, graphics, image and handwritten text. This method compresses the
scanned document based on blocked based pattern matching. Here scanned document is
first segmented into the blocks and then each block is processed separately. These blocks are
matched with dictionary block, if matched found then only the index of that block is used to
encode the original block otherwise parse the original block. In this way whole process is
carried out for each block and the whole document is compressed. The flags and dictionary
indexes generated are compressed with adaptive arithmetic coding. The quality of
reconstructed image is measured on the basis of certain performance parameter such as
Peak Signal to Noise Ratio, Mean Square Error, Compression Ratio etc.