https://github.com/neodoggy/lost-and-found-glasses
https://github.com/neodoggy/lost-and-found-glasses
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/neodoggy/lost-and-found-glasses
- Owner: NeoDoggy
- Created: 2024-12-25T18:36:04.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-12-25T18:49:35.000Z (6 months ago)
- Last Synced: 2024-12-25T19:27:30.085Z (6 months ago)
- Language: C
- Size: 0 Bytes
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Lost-And-Found-Glasses
Software has the model and the code for running the software
The firmware is for an esp32-s3 camera, feel free to upload it to your board with Arduino IDE
## Table of Contents
- [Training](#training)
- [Implementation](#implementation)
## Training
1. Navigate to the `./Software/yolo/` directory and open the `YOLO11.ipynb` notebook.
2. Start running from the `setup` cell and ensure all required `pip` packages are installed.
3. Run the `predict` cell to verify if `yolo11n.pt` is loaded successfully.
4. Locate and run the `import roboflow` cell:
- Provide your `api_key`, `workspace`, and `project` information to load the dataset from Roboflow.
5. Update the `data.yaml` file:
- Modify the `train`, `val`, and `test` paths from relative to absolute paths to avoid errors.
6. In the `Detection` cell, adjust the following parameters:
- `data`: Set to the absolute path of your `data.yaml` file.
- `epochs`: Specify the number of epochs as per your requirement.
- `imgsz`: Set the image size to your preference.
- Load the `yolo11n.pt` model.
7. Start the training process.
8. Once training is complete, locate the `best.pt` file and move it to the `./Software/model` directory.
## Implementation
**Note:** All implementation scripts must be run simultaneously.
1. Run `./Software/saveImg.py`:
- This script connects to the ESP32 camera at the URL `http://192.168.0.107/capture`.
- Captured images will be saved to the `./Software/images/input` directory.
2. Run `./Software/main.py`:
- Loads the `best.pt` model to predict the objects in the images saved in `./Software/images/input`.
- Displays recognition results using the `pygame` interface.
3. Run `./Software/STREAMmain.py`:
- If `main.py` detects an object, it returns parameters used to confirm whether an object has been scanned.
4. Run `./Software/monitor_db.py`:
- Continuously checks if the detected object's time is later than the `database last seen` time.
- If true, updates the database.
5. Run `./Software/showDB.py`:
- Displays all objects and their `last seen time` from the database using a `pygame` interface.