Lossless and lossy compression of images from the OMC experiment of integral project

OMC is Optical Monitoring Camera which is one of the experiments for INTEGR AL satellite. OMC is equipped with the CCD sensor with resolution 1024 x 10 24. Because of Low bits rate of Integral telemetry it is necessary to choos e interesting objects only and then compress the data of these subimages. M ethods of image compression have been very well described in the digital te levision (JPEG, MPEG) during last years, but these methods are not suitable for scientific purposes because of Loss of information. Lossless methods a re based mostly on the differential pulse coded modulation (DPCM) and reduc tion of entropy of differential images (RLE, Huffman, Arithmetic coding, et c.). Usual compress ratio is then between 2-3. Lossy algorithms mostly base d on integral transformation (Fourier, wavelet, Karhunen-Loeve etc.) allow us to compress images much more but these methods are not used in the astro nomical science because the definition of allowed Loss is not well defined. In the proposed paper the theoretical discussion and evaluation of some ap propriate lossless algorithms will be done. We will also try to define and discuss the criterium of possible loss for lossy algorithms.