School project that showcases use of REST API calls, trainer registration, mysql connectivity and API creation among other features to help scan pokemon cards or retrieve pokemon data.
0
stars
0
forks
source link
Tesseract is not able to recognize all characters, although we only need deck # and card # in this case #4
Tesseract library thus far works as it should, tweaking to the Optical Character Recognition (ORC) part needs to be tweaked some how.
Some research suggestions were to resize, crop and convert the image to grayscale.
Implementation of OpenCV was suggested for cropping but there's some issues currently with the dll used in windows and the dependency loaded. A possible version mismatch, or the use case of the current code in OpenCVUtils may be in compatible with the parameters used by this library.
Another error note: currently import java.awt.image.BufferedImage; is in charge of turning the image grayscale and in a compatible format with tesseract, however, OpenCV doesn't work with BufferedImages and a conversion was set in place. Somewhere in the conversion or after it is done and turned over for the cropping part, it fails.
Tesseract library thus far works as it should, tweaking to the Optical Character Recognition (ORC) part needs to be tweaked some how.
Some research suggestions were to resize, crop and convert the image to grayscale.
Implementation of OpenCV was suggested for cropping but there's some issues currently with the dll used in windows and the dependency loaded. A possible version mismatch, or the use case of the current code in OpenCVUtils may be in compatible with the parameters used by this library.
Another error note: currently import java.awt.image.BufferedImage; is in charge of turning the image grayscale and in a compatible format with tesseract, however, OpenCV doesn't work with BufferedImages and a conversion was set in place. Somewhere in the conversion or after it is done and turned over for the cropping part, it fails.