Novel context template design scheme for lossless color halftone image compression

Lossless halftone image compression using adaptive context template update

“…2 Then, the proposed method is compared with the line-based context template design method, 7 demonstrated in Table 2. Though 5 % of loss is observed in Airplane, the overall compression performance is comparable between the proposed method and [7]. Note that the line pattern-based template design addresses prescanning of whole image, prohibiting realtime implementation in electronic printing system.…”

“…Furthermore, the proposed template adaptation scheme is compared with a line-pattern preserving template design method. 7 In Table 1, the overall compression performance is illustrated between the proposed algorithm and the JBIG template. As Baboon image has many details and bi-level pixels are randomly distributed, we only save 20 % of bits compared with the JBIG.…”

“…[4][5][6] The JBIG2, for example, addresses 64 times larger memory size for context-based arithmetic coding than the JBIG. 2,4 In [3,7], prior to halftone image compression, input image is traversed to evaluate correlation between bi-level pixels. However, the prescanning of image for template design causes real-time implementation challenging.…”

“…Recently, a line pattern-based template design method was presented for a clustered dot ordered halftone image compression. 7 In context modeling, large-sized templates are often prohibitive for hardware implementation. [4][5][6] The JBIG2, for example, addresses 64 times larger memory size for context-based arithmetic coding than the JBIG.…”

“…[2][3][4][5][6][7] For example, in the standard JBIG, 10 bit-sized context template was used to reduce conditional entropy. 2 In [3], a template modification scheme was presented, which uses the autocorrelation of input image.…”