php - OCR not processing dot in image -


yesterday figured out tesseract ocr. i'm issue image.

before start, tested in 2 online ocr onlineocr.net , newocr.com , show output wanted. tesseract doesn't. (i don't know system site use)

the problem

please check image here: http://i.imgur.com/eorkxtg.jpg (please, don't embeded image).

using image, in both sites mentioned above, result:

/* output onlineocr.net , newocr have same */ @xxx john doe valor enviado    r$1,00 data     26/07/2017 01:46:09 status   aprovado transação: 4907115 <-- expected  

but in project using tesseract ocr, result. i'm using package:

““ªçãº: 49.913,15. <-- in project 

if see, image contain 2 dotted lines, , problem, edited photoshop , worked, image "order proof" (users can upload in system confirm order) , goal extract data instead of obligate users manually (date , transaction code "transação")

if need more info, let me know.

what using? i've tried?

i'm using homestead, , install tesseract ocr package (english , portuguese)

the way setup tesseract

$result = (new tesseractocr('image.jpg'))       ->lang('por')       ->psm(6)       ->run(); 


Comments

Popular posts from this blog

python - Selenium remoteWebDriver (& SauceLabs) Firefox moseMoveTo action exception -

html - How to custom Bootstrap grid height? -

transpose - Maple isnt executing function but prints function term -