Summary form only given. Bilevel image compression algorithms like JBIG, JBIG2-Generic, and PRECIS can exploit 1D or 2D peak autocorrelation in binary images like 'digital halftones', in order to achieve breakthrough boosts in additional compression. For hard to compress, but periodic halftones, boosts of factors of three or more times the compression ratios and similar increases in decompression speeds can be achieved (boosts defined v.s. the closest related, non-periodicity-exploiting algorithm (e.g. JBIG or JBIG2-Generic with AT>0 v.s. with AT=O, or PRECIS v.s. MMR)). Our peak autocorrelation finding method isolates the needed period for peak compression, two orders-of-magnitude faster than the prior art.

JBIG2 is a new standard for lossy and lossless compression of bi-level images. It exploits model-based coding for text and halftones; as well as nearby neighbor based coding for generic bi-level images. It can achieve compression ratios of up to three times those of existing standards for lossless textual image compression. JBIG2 also allows loss to be introduced while maintaining "visually lossless" quality for textual images, for which it can yield compression ratios of up to eight times those of existing standards. It produces even greater compression ratios when exploiting similarities between multiple pages and efficiently codes and integrates both scanned and generated bi-level images. JBIG2 was developed for applications having widely differing requirements such as facsimile, document storage and archiving, Web image coding, wireless data transmission, print spooling, and even teleconferencing. Accordingly, JBIG2 is structured as a "toolkit" of alternate capabilities to be selected based on the requirements of such applications.

Summary form only given. The Periodic Run Edge Compression for Image Systems (PRECIS) is a simple, fast algorithm to compress bitonal images containing periodic halftones. Its compression of periodic halftones doubles and often quadruples the compression obtained using the commonly used MMR algorithm. Its software execution time in compressing periodic halftones has been measured to be as short as half the execution time of MMR. In addition, one software embodiment of a simple version of PRECIS can be implemented simply by clever reuse of building blocks from an existing MMR implementation.

Lossless textual image compression based on pattern matching classically includes a "residue" coding step that refines an initially lossy reconstructed image to its lossless original form. This step is typically accomplished by arithmetically coding the predicted value for each lossless image pixel, based on the values of previously reconstructed nearby pixels in both the lossless image and its precursor lossy image. Our contribution describes background typical prediction (TPR-B), a fast method for residue coding based on "typical prediction" which permits the skipping of pixels to be arithmetically encoded; and non-symbol typical prediction (TPR-NS), an improved compression method for residue coding also based on "typical prediction". Experimental results are reported based on the residue coding method proposed in Howard's (see Proc. of '96 Data Compression Conf., Snowbird, Utah, p.210-19, 1996) SPM algorithm and the lossy images it generates when applied to eight CCITT bi-level test images. These results demonstrate that after lossy image coding, 88% of the lossless image pixels in the test set can be predicted using TPR-B and need not be residue coded at all. In terms of saved SPM arithmetic coding operations while residue coding, TPR-B achieves an average coding speed increase of 8 times. Using TPR-NS together with TPR-B increases the SPM residue coding compression ratios by an average of 11%.

Inspired by theoretical results on universal modeling, a general framework for sequential modeling of gray-scale images is proposed and applied to lossless compression. The model is based on stochastic complexity considerations and is implemented with a tree structure. It is efficiently estimated by a modification of the universal algorithm context. Several variants of the algorithm are described. The sequential, lossless compression schemes obtained when the context modeler is used with an arithmetic coder are tested with a representative set of gray-scale images. The compression ratios are compared with those obtained with state-of-the-art algorithms available in the literature, with the results of the comparison consistently favoring the proposed approach.

A JBIG compliant, quadtree based, lossless image compression algorithm is described. In terms of the number of arithmetic coding operations required to code an image, this algorithm is significantly faster than previous JBIG algorithm variations. Based on this criterion, our algorithm achieves an average speed increase of more than 9 times with only a 5% decrease in compression when tested on the eight CCITT bi-level test images and compared against the basic non-progressive JBIG algorithm. The fastest JBIG variation that we know of, using "PRES" resolution reduction and progressive buildup, achieved an average speed increase of less than 6 times with a 7% decrease in compression, under the same conditions.

This overview focuses on a comparison of lossless compression capabilities of the international standard algorithms for still image compression known as MH, MR, MMR, JBIG, and JPEC. Where the algorithms have parameters to select, these parameters have been carefully set to achieve maximal compression. Compression variations due to differences in data are illustrated and scaling of these compression results with spatial resolution or amplitude precision are explored. These algorithms are also summarized in terms of the compression technology they utilize, with further references given for precise technical details and the specific international standards involved.>