Taking a photo of a physical (printed) book page often results in an image with curved text. This makes text recognition much more unreliable.
The goal of this project is to build a routine that can recognize the curvature of the book page image and apply the unwarping (eg. CIWarpKernel) based on that information. The output of the routine would be an unwarped image with straightened text lines.
6 freelancers are bidding on average $1079 for this job
Hi I have read your description carefully. I have rich expereince with image processing and OCR. I think I can help you with my skill and expereince Hope your kind contacts. thanks.