how i can compress double data type of matrix without loss any information?

i have a double data type matrix (image) i want to compress it without any loss of information with keep the data type of image (double).

[Walter Roberson]

In the general case, you cannot be sure that you will be able to reduce the amount of memory being used.

[Iain]

Double format (15 significant figures and a range of 10^-308 to 10^+308) numbers are excessive for most normal images.

Single format numbers are still excessive. (7 significant figures and a range of around 10^-38 to 10^+38)

16 bit integers compare extremely well, and this corresponds to the "raw" format of many relatively high end SLR digital cameras.

8 bit integers, again, compare well. This is what your normal PC monitor can handle. (8 bit Red, 8 bit Green, 8 bit blue [8 bit transparency etc.])

Provided that your numbers are sensibly scaled (0 to 65535), you really won't need anything beyond 16 bit integers.