This program generates recipes based off a picture of a receipt
Progress so far:
Natural Language Processing
First, I preprocess the image to prepare it for analysis by converting the image to grayscale using cv2. Then using an OCR (Optical Character Recognition) tool call pytesseract, I extracted text from the image and tokenized it to identify the names of food ingredients and filter out any non-food items that may have been included. My dictionary is small at the moment and needs to be improved.
The requests library to send an HTTP request to the website's search API and parse the resulting HTML or JSON data to extract the search results.
Plans for the future:
Creating a front-end interface with Kivy
~ Allows the user to take photos of multiple receipts to add to personal ingredients list
~ Allows manual addition to the ingredients list
~ Allows deletion from personal ingredients list
Extract and reformat search results into app