Persian number can't detect well

[arzmaster]

Hey guys I have weird problem with easyocr .Im going to extract these persian number from images for example for this image easyOCR should extract '١٤٠١٢٢٠٣٧٩٢٤۶٩٣'

I found out easyOCR detect my numbers as 2-3 groups of numbers and add extra space between numbers and changed order of number

I copied and pasted the exact number result here (١٢٢٠٣٧٩٢٤٥٩٣ ١٤٠) , if you noticed and check this number by hexeditor realize its not correct number , its ( ١٢٢٠٣٧٩٢٤۶٩٣ + space + ١٤٠ ) but in output python showed its as different

when I paste the result in vscode environment there is this error Invalid character "\u661" in tokenPylance

how can I fix this weird error ?

[rkcosmos]

It has to do with ' ' space characters. Whenever EasyOCR see space, it will rearrange reading direction to left-to-right. The possible fix is to write a python script to detect space in output, and rearrange the text back to right-to-left.

[ftmasadi]

Hey guys I have weird problem with easyocr .Im going to extract these persian number from images for example for this image easyOCR should extract '١٤٠١٢٢٠٣٧٩٢٤۶٩٣'

I found out easyOCR detect my numbers as 2-3 groups of numbers and add extra space between numbers and changed order of number

I copied and pasted the exact number result here (١٢٢٠٣٧٩٢٤٥٩٣ ١٤٠) , if you noticed and check this number by hexeditor realize its not correct number , its ( ١٢٢٠٣٧٩٢٤۶٩٣ + space + ١٤٠ ) but in output python showed its as different

when I paste the result in vscode environment there is this error Invalid character "\u661" in tokenPylance

how can I fix this weird error ?

HI I have work on this field how contact you?