How do I avoid the tesseract error of too few characters when trying to perform tesseract osd on PIL/opencv processed image?
I am trying to perform tesseract osd on a file that has been rotated by an angle that is not either 90 or 180. The crux is, after rotating the image by an angle(!=90,180,270) and then rotating back by the same angle using either PIL or opencv, I am unable to perform osd because tesseract throws “too few characters” error. If I take screenshot, and perform osd on the screenshoted image osd work fine. I am wondering why tesseract isn’t able detect orientation after processing the image in python. Is there a way I can get around this issue?coderest of the code