Traint Tesseract version 4 to identify a font. And supply the files and syntax to use the trained data for OCR.
Tesseract is capable of recognize 99% of the strings without any training, after rescaling and Grayscale with ImageMagick.
But it needs to be better. Perferably without ImageMagick
Please confirm that You have understood that it is Tessercat version 4 !
I have attached short example.
7 pekerja bebas membida secara purata $214 untuk pekerjaan ini
Though I am new here but my team has 7 years of experience into OCR,Tesseract all versions. Can very well execute this Project as the team has good hands-on experience.