- Python >= 3.6
- Apex Legends
- Tesseract
$ git clone git@github.com:Rosswell/apex-stats-ocr.git
$ cd apex-stats-ocr
$ pip install -r requirements.txt --user
- Change the name of the csv file to your liking
- Adjust the
stats_filevariable, if necessary.stats_fileis the name of the file you'd like to write stats to, and must be in the same directory asapex_ocr.py. - Adjust the
monvariable, if necessary. The default value is1920x1080. This is a representation of the top half of your monitor, because image processing takes less time for smaller images, and the stats on the Match Summary screen are on the top half. Thus, themonvariable for a1920x1080monitor should be(0, 0, 1920, 1080 / 2)(which is the default value).
NOTE: You must run Apex Legends in windowed mode for this to work. This is honestly just because I couldn't figure out how to capture full screen mode, and Bordered Windowless is good enough.
Simply run
$ python apex_ocr.py
Which should print Watching Screen... to the console, to indicate that the program is actively monitoring screenshots
for the Match Summary screen.
A higher-pitched beep during the Match Summary screen indicates that the Match Summary screen has been recognized. A subsequent lower-pitched beep indicates that the requisite duplicate images have been taken, and you can now navigate away from the Match Summary screen. The amount of time between these two beeps should be under 5 seconds (about 2 during my testing, YMMV). The OCR processing happens subsequently, and successful writing to the stats file is indicated by a console message.
I legit have no idea what I'm going when it comes to OCR, so if you can get a more consistent rate of interpretation of the screen, please go ahead and submit a PR. Also just generally I don't anticipate many people using this, so if you have any other contributions in mind, do the same, as this is currently very rough.
Apex-ocr is distributed under the terms of MIT license.
See LICENSE for details.