0

I'm starting to use Tesseract, my goals is to be able to read basketball game images and read the scores and the game time. I have attached a sample image. Tesseract was able to pickup some, this is what it detected for the attached image.

‘\  semis 
1ST 50.9

Some image, it's not able to pickup anything. I went thru the training section but it talks about fonts and such which I do not know. Different games may have a slightly different scorebox. Does anyone know how I can train Tesseract to be able to pickup the boxscore data of the screen or is there a different solution out there?

thanks testimage

Thaison Nguyen
  • 131
  • 1
  • 10
  • Do you run tessecract on all the image or just on the score box? if not try to first crop the score box and run tessercat only on it after doing some preprocessing like thresholding etc.. . If this gives you a good result try find the rectangle that as the score from the original image using something like http://stackoverflow.com/questions/1817442/how-to-recognize-rectangles-in-this-image – Amitay Nachmani Mar 02 '17 at 09:16
  • @AmitayNachmani I ran Tesseract on the whole image right now. You're recommending extracting the scorebox out on it's own image and run Tesseract? Is there a way to do this efficiently? as Different games have different location where the scorebox are at and they all look different. thanks – Thaison Nguyen Mar 02 '17 at 15:27
  • @Tienkamp I care about the results but I will be processing thousands of these, hence using a web api might be a little expensive. Also, I tried ocr.space and I get this for the image, "BON us BONUS " which still missing vital information on it. – Thaison Nguyen Mar 02 '17 at 15:30
  • Yes i recommend to extract the scorebox. You can try just insert it like this but i think you will get better results if you convert to gray scale and threshold the image – Amitay Nachmani Mar 02 '17 at 16:16
  • @AmitayNachmani thanks, do you have any recommendation on how to extract this? Since every network/games are slightly different, some my be top corner some bottom. I was hoping I can somehow company maybe 5 frame and spot it out but google search, I can't find anything. Maybe i'm using wrong keywords in my search – Thaison Nguyen Mar 02 '17 at 18:06
  • try to read about rectangle detection in image processing – Amitay Nachmani Mar 02 '17 at 18:30
  • @AmitayNachmani thanks, reading openCV, will give it a try tonight. – Thaison Nguyen Mar 03 '17 at 13:56

0 Answers0