Releases: rad10/OCR-to-CSV
Documentation Explosion
Release Conventions
1.6.2 Release
This release comes with a whole change on the code. Rather than have the entire program in one file, the code has been split into multiple files for readability and ease in development. Thanks to these changes, the scanner has been made more efficient, requiring less time to scan an entire signin sheet while also being more accurate than previous versions.
This release comes with these notable changes:
- new tesseract scanning method (#12)
- everything is placed into specific modules for their subjects (#7)
- GUI has been improved upon visually (#5)
- Sanitization options have been included (#10)
- better organized scraper, including dates (#16)
This has created the following improvements - Significantly higher accuracy (#11)
- Faster runtimes (as it requires less images scanned) (#6)
Modularization
A Whole rework on the program
- Made module files to hold many functions
- Redid scanning function to use hOCR output instead of string output to gather alternatives
- runs faster and more accurate than previous versions.
- is missing sanity checking mechanisms for more accurate results. That is on the way.
A Better debugger
This version now better deals with debugging for developers. It also has better quality as empty lines will no longer show up. Time reading has also improved.
Prelease fix
Sadly, I pushed out the last version without proper testing on computers not initially setup with the software. Issues have been fixed, and running the program for the first time should be much easier than before.
The first release
This is not the final code, but it is in a state that allows it to do its job. It doesnt do its job as well as is could do, but that is for later versions.