hi all,

under windows os or directx 32/64 api,

what are the recommended methods to grab screen data, parsing (assume those are images) continuously for 6 hours daily?

assume there are about 80% of the whole screen 1920 x 1080 pixels we need to process every second or two seconds.

any ideas?

* something cool, https://github.com/tesseract-ocr/tesseract
might need to test how many small rectangle and how many count of png file it could process in a second.

* now how to grab screen in memory and pass them to tesseract

* using office sharepoint 2007 (free) MODI to ocr from tiff
https://community.spiceworks.com/how_to/30155-how-to-deploy-modi-automatically-without-ms-office-licenses-or-media

Why are you grabbing screen data? Rather than going through all the steps to transfer the images to screen and back again into main memory, why not just do it directly without the intermediate stage?

Why are you OCRing? Rather than rendering the characters to pixels and back again into characters, why not just take the data directly without the intermediate stage?

Why are you storing in a database? Rather than taking data from a database, rendering to pixels, and sending to screen memory, and then doing the whole process in reverse, why not just take the database directly without the intermediate stages?

hi revolution,

there are 2 sets of data, first one output by active x control in ie browser, second one output by java applet,

i am thinking about OCRing to solve grab data from above 2 places in one time, i am not really sure how render pixels back to characters, and how about anti-aliased pixels, not really sure how to reverse them,

i need to store those OCR output in database and send them out through telegram bot api.