Repository for a school group project for a machine vision class (VSV).
Choose a camera and light setup and create a software solution to indetify defect cables.
Solution is a neural network created on images taken in the school laboratory. The dataset was expanded and then labeled using RoboFlow.
Neureal network itself was also used using Roboflow utilising ImageNet as a starting point
Final solutions is a predict.py script that provides a GUI for user to appload a photo and then it uses Roboflow's API to connect to our workspace and predicts the cables state using our network
Warnig API key is not included in the github code
- customtkinter for creating the GUI
-
pip install customtkinter
-
- Roboflow for connecting to the roboflow API
-
pip install roboflow
-
- API Key needs to be placed into the predict.py script as a string
-
app = App("!!API_KEY_HERE!!")
-
- Run the script