OCR Image Quality

[USMCGalloway]

I have the UDF function _OCR working with the example .bmp file loaded by PingPong, but I am having trouble getting my font into the program. I think it has something to do with the way I am grabbing the image. At this point I am using  "Snipping Tool" but have also used print screen to try to capture the image of the numbers I am trying to grab.

Before X800

Capture.bmp

After X800

Capture1.bmp

 

I am not sure how to go about coding for the variation in color between the row/columns since they are not a flat black(0x000000), rather they range from 763336-000000 ish. I know with pixelsearch you can have a buffer zone to include shades within a percentage of a color you are really looking for, is this something that would apply here as well? 

I have tried a few other locations  within the citrix app, but I have the same problem with the image being to pix-elated, and the other alternative takes to long to navigate too.  

Any help is much appreciated.