In this chapter, we created a new application named Literacy. In this application, we recognized characters on images using the Tesseract library. For images that have well-typeset characters, Tesseract worked well; but for the characters in photos of ordinary everyday life, it failed to recognize them. To fix this issue, we resorted to the EAST model with OpenCV. With a pretrained EAST model, we first detected the text areas in photos and then instructed the Tesseract library to only recognize the characters in the detected regions. At this point, Tesseract performed well again. In the last section, we learned how to grab the desktop as an image and how to select a region on it by dragging the mouse.
We used several pretrained neural network models in this and previous chapters. In the next chapter, we will learn more about them; for instance, how to use pretrained classifiers...
