Computer Vision News - November 2019

Summary We Tried for You 26 Now when running this code on our test case we get the correct answer: Another feature of Tesseract is the extraction of location (i.e. bounding box) for each character. To do this, we can use the image_to_boxes method. The following code extracts the bounding boxes for each character and draw it on the image. You can add it to the main function we introduced earlier. The result of the above code looks like this As you can see the characters bounding boxes are quite accurate on our test case, but this is not always the case. Let's see some limitations of the Tesseract.