Hi, ankush,
I'm trying to use your project to generate some ocr training images and i found your code only support opencv2
Since now opencv3 is the popular version (pip does not provide opencv2), so i transform cv2 code to cv3.
A requirement file and readme is updated