I will be documenting my journey developing my Grocery Tracking application. This will be for a form of accountability as well as motivation to continue the development effort.
Vision:
It is difficult to keep track of all the groceries that you buy each week. Often, food can go to waste if you do not optimise the food to eat before expiry. Tracking the food purchases on a weekly basis, also tracks what items are part of your core diet, and could make future shopping much more quicker, which is a feature that online shopping has now, that has not been reflected into physical shopping.
As a future goal, there could be a rating system of each shop. At the end of the week, can also rate how successful you were in consuming all of the food.
An ambitious goal, would be integrating with some third party application (if exists) that can add the shopping order, and output what potential recipes could be made.
Goals:
- Set up android studio
- Set up camera capture of receipt
- Connect to Firebase ML
- Table of receipt results
- Set up Database
- Improve Firebase ML for reading receipt
- User profiles
- Third party app integration
- Animations
- Clean UI
Summary:
I have since put this project on hold, and will pick this up again once I gain further knowledge in ML for text recognition. As it seems that the Firebase ML library for OCR, is not accurate enough for this intended use case, or the quality of receipt text is difficult to train.
- It may be that I create and train my own ML library.
Logs:
- August 20th 2021
- Connected Firebase Database connection
- Not happy with OCR accuracy.
- July 25th 2021
- Spending time learning how to build a well rounded android application
- Struggling with Firebase Database connection
- July 28th 2021
- Documenting wireframes and vision for the project.
- Aim to learn the resources before going again
Wireframes:
