/ApexReader

Analyze screenshots Apex Legends game summaries using Google's Tesseract OCR

Primary LanguagePython

ApexReader

Analyze screenshots Apex Legends game summaries using Google's Tesseract OCR

Details

Uses Google's Tesseract OCR in conjuction with template matching via the opencv-python library. Parsing a screenshot follows this general pattern:

  1. Image is loaded from filepath via the cv2 module and converted to grayscale
  2. Loaded image undergoes thresholding to aid in parsing
  3. When a specific attribute is requested to be parsed, the appropriate template image is selected
  4. The selected template is found on screen using template matching, returning coordinates of the template image
  5. The location of the template is used to determine the region onscreen of the desired attribute information
  6. A cropped image containing the attribute information is finally passed to the OCR module
  7. The result of the OCR processing is saved for the attribute

Additionally, it's worth noting that attributed aren't calculated until requested. This allows for the ApexReader module to feel more responsive, as the entire image isn't parsed upon loading an image. This is done by using the @property decorator to allow properties to be calculated on their first use, instead of on initialization.

Example usage:

Let's extract the information from the following screenshot

We can do this simply with the following script

from ApexReader import ApexReader
from pprint import pprint

ar = ApexReader()
ar.load(r'.\test_imgs\img4.png') # Or whatever screenshot you want to analyze

pprint(ar.game_data)

Which returns the output

{'num_players': 3,
 'placement': '1',
 'players': {'player0': {'damage': '2250',
                         'kills': '12',
                         'name': 'lfegirltocapture',
                         'respawns': '0',
                         'revives': '1',
                         'survival_time': '16:55'},
             'player1': {'damage': '2799',
                         'kills': '10',
                         'name': 'OGRealStimJimmy',
                         'respawns': '1',
                         'revives': '1',
                         'survival_time': '16:55'},
             'player2': {'damage': '982',
                         'kills': '4',
                         'name': 'syuraponda 4',
                         'respawns': '0',
                         'revives': '0',
                         'survival_time': '1855'}},
 'squad_kills': '26'}
>>>

We can see the screenshot isn't read perfectly but it is very close, too which we can thank Google's Tesseract OCR.