A higher and better precision can been achieved while using more recent observation techniques and superior detection
sensors. It has brought a very rapid increase of data amount as a result. The high spatial resolution (up to ten million
pixels) and a high bit level grey scale images (quantization depth up to 16 bits) are used in astronomy and other scientific
applications. A very large volume of image data has been taken during operation of a modern project of automatic (i.e.
robotic) sky observation systems. The coder based on the Karhunen-Loeve transform (KLT) has been chosen for
astronomical image compression in this paper. The astrometry and photometry measurements have confirmed a
possibility of the coder blocks arrangement with production of an accepted error and a sophisticated data stream.