InjectStrap

This repository is designed for extracting features from videos as part of the Smart4ALL experiment. The experiment involves using thermal videos to extract thermal properties and generation cycles of components from the Elvez compnay. The videos were recorded using a PI thermal camera. This repository can work real-time alongside a camera.

In the video that follows, a component can be observed as it moves in front of the camera. By studying videos like this one, we can record the thermal properties of the component, as it appears in front of the camera, in tabular form. The brought videos is just a sample.

sample-video.mp4

After identifying a static frame that encompasses all parts of the component, we gather information from 11 distinct regions of the component. These regions are illustrated in the image below.

How to run the program

Running this program is straightforward. To generate a dataset from the videos, the main.py script must be executed. This script accepts two arguments: vid_dir which is the directory containing the input videos, and save_dir which is the directory and name of the generated dataset. By executing the following code lines, the script will run and the generated dataset will be saved in the specified directory and filename.

 python main.py --vid_dir "./vid.ravi" --save_dir "./result/dataset-vid1.csv"

or

 python3 main.py --vid_dir "./vid.ravi" --save_dir "./result/dataset-vid1.csv"

Requirements

  python>=3.8
  numpy>=1.23
  pandas>=1.3
  opencv-python>=4.6

Acknowledgements

This Knowledge Transfer Experiment has received funding from the European Union’s Horizon 2020 research and innovation programme under the SMART4ALL grant agreement no 872614.