SmartBin AI
I am making SmartBin AI, which is a project that can analyse both the type and percentage of food waste occurring at a school campus/canteen/kitchen. This project will use a Raspberry Pi, as well as its Module 2 camera to capture the image, after taking photos, it will run a Region of Interest (ROI) analysis on all images to find out which photos has plates. After receiving such list of images, it will pass all of them through the OpenAI o4 model and find out what kind of food it is, as well as the percentage of waste.
Followers
Abū al-Barāʾ
Check their projects out: Moaz Website generator, Image Compressor, Moaz Video Downloader, Typing Speed Test, QR Code Scanner, Python Learning Website, QR Code Generator, Text Transformer, Hangman Code Generator, Text to voice converter, Currency Convertor , Pac Man Game, Text to Handwriting Converter, Beautiful Chess
Ship Your Project
Get ready!
Once you ship this you can't edit the description of the project, but you'll be able to add more devlogs and re-ship it as you add new features!
Timeline
Made just the basic landing page of the dasboard tab for the website. Also got the datatable sorted out on the backend side of things for this.
I finally just finished! I configured the Raspberry Pi OS to also show my app in the desktop as well as the side bar
I completed the combining the code as well as the UI interface for the COMPLETE app. this means that the user can simply use it as if it was a web app they were launching! i made this using tkinker and python, really happy with this. i also managed to make the colors match with my color scheme, etc, and all.
Finally completed the script that constantly takes images every 0.25 seconds and saves it if it satisfy the Region of Interest (ROI) function for plate detection. Now I am going to focus on developing the Graphical User Interface (GUI) for the project...
Completed making the model where I am connecting my python code with the gpt-4o model specifically for image classification. moreover, I created a detailed prompt.txt(more than 200 lines of code) to make sure the model will output data in the exact format I need it as.
Next I need to make the code that will actually take continuous photos from the raspberry pi and use ROI(region of interest) to sort the helpful pictures from the unhelpful ones..
Took quite long to complete the next part but I got it done!!! I managed to change the UI interface from terminal or command prompt to a Graphical User Interface using the tkinker python library.
The general objective of this was to make the code more user friendly, especially for someone who isn't technical.
My next plan is to develop the code that takes continuous images and sorts them depending on which has a plate and which one doesn't.
I completed the part of the code that takes images from a folder and converts them into base64. Since computers cant process images, it requires the image in base64 format, which is a text format for the computer to understand the picture. I have also managed to pass the images through to the openai model, however I am receiving some errors, that I will troubleshoot later on..
