Finish a number recognition software, Nearly done, just a few tweaks and features

The aim of this whole software is to update a database with the latest number found in a list.
If at any point you feel I have made a mistake or the code can be improved in a different way other than I have mentioned then go with it.


I have this in two parts, the GUI part which at the moment is used to define the title area to look for which then points to the list of number data to capture.
When this information is selected I then run the OCR code which previously has worked just fine.. within another setup.

When I now run the OCR part I am getting errors. (not recognising numbers etc) (angle data not found or title area not found ..)
---Update, the error describes that the angle.dat file is not completed, so adding more angles or something similar should be done to fix this.---


There is a seperate application

So that is what needs to be repaired/ fixed might well be simple...
I would like the application to be fixed to the point it can capture the numbers and ensure it is capturing data accurately.
-----Update, this may include configuring Opencv/ OCRtesseract with an updated method to improve the accuracy of number reading/ changing colour/ contrasts etc-----
If it is easier, converting the OCR engine for a more accurate version may be done....
although I have asked for a simple feature to train tesseract data to improve the accuracy of number reading further down this document.









FEATURES TO ADD IN EXISTING GUI

I would like an edit in the software gui to define a MySQL database and server to communicate to when highlighting each numbers area and be able to save the setup as a preset/ configuration file.
-1- When highlighting each area title, define which database server and table to send the data to
-2- Define which workspace this instance will be running on
-3- Add button to run the next application which is the OCR part..(at the moment a new terminal is loaded under the OCR application directory)
-4- Ability to save table and selected MySQL database server details as a setup file,
---------CAPTURE SCREEN ON MULTIPLE WORKSPACES (a system tweak)----------




FEATURES TO ADD TO MAIN C++ APPLICATION CODE - (Ignoring certain things..)

^^^^^ I would like a feature to make sure when the capture area is moved on the screen, it does not affect the operation of the accuracy of the program. ^^^^^
-1- Detect each time the number list movement/ has the list changed direction?/ has the font changed, no output is needed as long as the numbers update accurately.?
*note it is a title area that then point to list of numbers that update and move around the screen periodically
-2- Compare latest collected numbers with the database to make sure the correct area is being captured. (IE if the location title changes for the numbers but the list of numbers has not..)
This can be done by checking the last five captured numbers with the database and compare them to the current found list of numbers captured.
-3- Add ability so switching between workspaces does not break any operations.



ADD COMBINE MODE

This will most likely be just two/three or four seperate workspaces on the same machine running an instance of the software.

-1- Detect if another instance is capturing the same area (area title) or and compare captured data for accuracy
-2- Add tickboxes to determine how many of the running instances should be correct to allow an update to the mysql database to happen
-3- Only update mysql if selected amount of instances captured lists are the same.




ADD EASY TO USE GUI INTERFACE TO TRAIN NEW NUMBER RECOGNITION DATA SETS WITH TESSERACT TO IMPROVE/CORRECT OR TRAIN NEW NUMBER DATA

-1- click train number, then highlight area with number image, type what number is, add it to the trained data.
* note - - there is a command in tesseract to train number data with.







Extra notes: My eyesight is bad so I am trying to make it easier for me to read onscreen information.

slowly by adding features to this software I aim to make it easier to use a computer...

Thanks for anyone reading