MULTI-COMPONENT STILL IMAGE COMPRESSION Contact Kennith Whitson, Central Imagery Office, 8401 Old Courthouse Road, Vienna VA 22182-3280, (703) 275-5649, FAX 703-275-5088. The Central Imagery Office (CIO) seeks candidates for a new multi-component still image compression algorithm for inclusion in the National Imagery Transmission Format Standard (NITFS). Algorithms are sought for compression of 8 to 12 bit multi-component imagery to rates below 0.25 bits per pixel per band (bpppb), with quality loss at 0.25 bpppb of no more than 0.5 on the National Imagery Interpretability Rating Scale (NIIRS). The multi-component imagery class includes all images which contain more than one band of image information. Multi-component images are often composed of bands which portray different spectral representations of the same scene, as is the case with color and multi-spectral images, however, many images composed of several non-spectral bands are also included in this class. For instance, time sequences, stereo pairs, sequential spatial slices and MRI images are also considered multi-component imagery. Interested parties must respond directly to the POC identified above by 043095, expressing their interest and giving a brief description of their algorithm. Non-government evaluation assistance will be provided to the CIO, therefore, proprietary information should not be disclosed. Offerors must be willing to relinquish or sell to the government at a reasonable price all proprietary elements of the algorithm for inclusion in the NITFS. Respondents whose proposed algorithms appear to be consistent with the stated requirements will receive an evaluation packet consisting of approximately four original images and sample reconstructions for each of these images to be used for self-evaluation. Offerors whose algorithms cannot perform better than the example reconstructions should not continue with the evaluation. The evaluation packet will also contain a detailed questionnaire for offerors about their algorithms, and detailed instructions for further stages of evaluation. The algorithms will be assessed on reconstructed image quality, compliance with existing standards, cost to the government of acquiring rights to the algorithm, portability to a wide variety of platforms, computational complexity, bitrate control, and vulnerability to bit errors. The government reserves the right to accept any number of algorithms for inclusion in the NITFS, or not to accept any proposed algorithm. The CIO will not pay any costs associated with responding to this announcement. No contract award is anticipated as a consequence of this announcement.

Loren Data Corp. http://www.ld.com (SYN# 0543 19950331\SP-0001.MSC)