The simpliest technique is to measure angles between different faces of the crystal by goniometer of optical microscope. If the crystal has cubic symmetry you can use lightfigure technique for roughly evaluation face indices. For precise identification X-ray crystallography techniques for single crystalls is better to use.

If you don't know the symmetry of the crystal you can use the old technique described by Alexandra. In case your crystals are big enough and has a sufficient number of planes there are good chances to derive symmetry as well as indexing since all planes have to be describable by small integers, i.e. commonly up to 3 or 4, however his is only a rule and exceptions may exist, especially for complex structures.

For me it sounds that your crystals are perhaps very small so that you cannot use and optical goniometer to measure the planes by reflection. In case you only have small crystals visible in electron microscope the entire business becomes more challenging, but it is still possible, although more complicated. You can use then any single crystal structure technique, like XRD, SAED or EBSD. The biggest problem then will be the indexing of the planes since the diffraction does not work on the visible planes but on the volume of the crystal. The challenging part is then to find ways to assign the discovered indexings of interferences to specifically visible planes. I would highly recommend to post one of your images. Sometimes (especially cubic) phases already display very easily recognizable habits so that every mineralogist or crystallographer is able to identify the visible habit.    

use mercury software to identify the crystal faces. But it needs crystallographic Information file as a source.

Dear T. Karunanithi: There are hundrets of calcite forms known, all with the same structure! How do you think a software can derive the habit of a crystal using a CIF file? 

You should take classical method from XIX century. You should follow exact procedure described in textbooks on crystallography.

 @ V. Rakin: I guess that the crystal he has are a) not of brilliant quality regarding the reflectivity of planes, and b) simply too small in order to fix them on a goniometer. But in fact you are right. The major work has been done in the second half of the 19th century and the very beginning of the 20th, e.g.  by Wulff. I am not sure whether students are still tought in measuring crystals on a goniometer (except perhaps in Russia).

For your purposes will approach not two-circle goniometer, but parabolic in which simply enough to establish symmetry and to identify the faces of a small crystal (>0.2 mm) .