1. If ImageMagick is not installed then install it. After that install ?Cuneiform port.

2. After package installation login as ordinary user and run script ocr:

qocr rus

It will create in your home directory qocr directory and next subdirectories for each language:

ls /home/your_login/qocr

bul  dan  dut  est  ger  ita  lav  pol	rum  ruseng  spa  swe  txt  uzb
cze  dig  eng  fra  hun  kaz  lit  por	rus  slo     srp  tur  ukr

The "txt" subdirectory is used as output directory for recognized OCR text. 
 
3. Prepare images - Scan images it with 400 DPI and open each in GIMP. 
Rotate them vertically, cut on blocks and frames save each in separate file. 
Then add Brightness/Contrast to clean of dirty dots and old paper greyness for each  block. 
Save all images. If ImageMagick is not installed the Cuneiform could recognize only *.bmp images. 

4. Place all images to directory with appropriate lang. For example if you have russian lang. images then copy them to ~/qocr/rus directory, and if you have english lang. images then copy them to ~/qocr/eng directory. If you have mixed mostly Cyrillic ans less English scanned imaged place them to the ~/qocr/ruseng directory. 

5. Then run ocr script with parameter of lang. to OCR text.

qocr lang

For example, if you copy images to ~/qocr/rus directory then run 

qocr rus

And If if you have english lang. images in  ~/qocr/eng then run:

qocr eng

If while recognition you will see next errors:

Assertion failed: 0 file /usr/share/cuneiform/cuneiform_src/Kern/lns32/src/sbambuk.cpp, line 343

Press <Space> to continue execution, <Esc> to abort

Then simply press "Space" and then "ENTER".
 
6. When OCR finished open in ~/ocr/txt/ text file text-packet-*.txt
and check it by spellchecker program. Delete all images.

7. That's all!