Text image on multiple lines


Is there a way for EggPlant’s OCR to detect a text image of a long sentence input that overruns several lines (word wrapping)? I cannot parse the sentence since I don’t know how it will be displayed on the image. I need to get the coordinates values of the sentence.

Thanks for your help.

No, the OCR can read multiple lines, but it can’t be used to find text spanning multiple lines, even if you know specifically where the line breaks occur. The best workaround is to look for a number of words – preferably unique on the page to that string – and use those locations to determine the location of the string. You can use the smallest and largest (x,y) values for the found images to create a rectangle.