I have an image of a PDF containing several text fields that I want to pass to AWS Textract for text extraction. I have thus created cropped images of these specific text boxes. In order to keep all of these boxes straight (and able to determine which image came from which field), I am operating primarily within a dictionary and created a function to call on the Textract API and enter the resulting response as an entry in the dictionary. An example of the dictionary and the code is below.