This repository contains a GUI program that can:
- Load thermal videos captured using FLIR camera.
- User can create an oval on first frame of the video, containing region of interest
- The program uses a point tracking model (tapnet) to track the oval throughout the video
- It computes average values within RoI throughout the video and user can export it as excel file, and a scatter plot.
Download the FLIR SDK from the official website. After visiting the website, you have to:
- First select SDK (Software Development Kits)
- Then select FLIR Science file SDK
- Then install the SDK for the windows.
Windows users: Once you download the SDK file, there is a .exe file inside the compressed file you downloaded. Install it but remember the path you choose. After installation, if you choose the default path, then you can go to C:\Program Files\FLIR Systems\sdks\file\python to see where you can install SDK for python.
For installing it, you have to run the command prompt as adiministrator from the search bar menu:
Then you have to change directory to where SDK is installed and install the SDK:
cd C:\Program Files\FLIR Systems\sdks\file\python
python setup.py install
Note: For the code above to work, it is expected that Microsoft Visual C++ is installed on computer. If not:
- Download it from the official website by clicking on Download Build Tools button.
- When installing, make sure Desktop Development with C++ is activated and do not change anything from the installation details:
After installation of SDKs, install the dependancies related to python program by installing the packages with the command below inside the project directory:
pip install -r requirements.txt
pip install -r tapnet/requirements_inference.txt
Also install the torch if you haven't already based on the instructions provided in official website.
Finally, download the checkpoints for point tracking from here and place it under tapnet/checkpoints directory.
python main.py
In case you have a strong GPU and want to process faster, you can increase the batch size (default is 16):
python main.py --batch-size 64