The purpose of this project is to create a quick and lightweight script to classify images of an SLS process. For this purpose a vgg16 model was trained and saved.
The contained script starts a scheduler which calls three functinos every 35 sec:
- load latest modified image from folder.
- use vgg16 model to predict if it contains curling or not.
- still missing is the conversion of the raw images to the rgb jpeg format.
- set up a virtual environment with anaconda and the contained yaml file
- start script in context of environment either in the terminal or through pycharm