I'm totally new with Tesseract and its usage and I tried several hours to
get a suitable result for the attached web-cam picture of my gas meter. I
do use tesseract with the PHP wrapper from
here https://github.com/thiagoalessio/tesseract-ocr-for-php and
installation and recognition with the gi
r($thumb, IMG_FILTER_GRAYSCALE);
imagefilter($thumb, IMG_FILTER_BRIGHTNESS, +50);
imagefilter($thumb, IMG_FILTER_CONTRAST, -50);
As a result I got at least a result: 39984
Its still not correct, but better than no identification of any letter...
How can I improve?
Martina Ilgon schrieb am Montag,
2 matches
Mail list logo
