Welcome to the world of automated health screening with AiThermometer! This nifty piece of software allows you to harness the power of thermal cameras to efficiently measure the temperature of individuals. It’s especially useful for early screening of fever symptoms. Let’s dive into how to set it all up and troubleshoot common issues.
Getting Started with AiThermometer
AiThermometer is developed using Python3 and has been tested on Ubuntu 18.04 with NVidia drivers, Cuda 10.0, and Cudnn 7.6.5. You can run it in a Docker container, which we will explore later.
Installation Steps
To get AiThermometer up and running, you need to follow these installation steps:
- Install the Requirements: Before anything else, make sure you have the necessary packages.
- OpenPose: Follow the instructions in the OpenPose repository to install it in the ai_thermometer/openpose folder.
- Flir Image Extractor: Install using the command
pip3 install flirimageextractor
. - OpenCV: Run
apt-get install python3-opencv
followed bypip3 install opencv-python
. - Install Docker: If you prefer running AiThermometer in Docker, follow the instructions in
docker_howto.txt
to install Docker.
Using AiThermometer
Once set up, you can run the software with either radiometric or non-radiometric images. Here’s how:
With Radiometric Images
If you have access to radiometric images, you can execute:
python ai_thermometer.py --image_in ./samples/IR_2412.jpg --image_out ./results/IR_2412_out.jpg --radiometric True
Without Radiometric Images
If you are using uncalibrated thermal cameras, you need to provide a reference pixel and temperature. For example:
python ai_thermometer.py --image_in ./samples/image1.jpeg --image_out ./results/image1_out.jpg --reference_px 200 --reference_py 400 --reference_temperature 31
Understanding the Process: An Analogy
Imagine you are a chef in a kitchen. You have a recipe (the code) that requires precise measurements (the camera readings) to create an exquisite dish (temperature check implementation). As a chef, you gather your ingredients (installing required packages), prepare the workspace (setting up your environment), and follow the steps outlined in your recipe (running the program). Just like tasting the dish along the way to ensure it has the right flavor, you can take various measurements along the way to verify that your results are accurate. Each component works in harmony to achieve a perfectly cooked meal—or in this case, an accurate temperature reading.
Troubleshooting Common Issues
If you encounter problems while using AiThermometer, here are some troubleshooting tips:
- Ensure that all packages are correctly installed and paths are set.
- Check that the reference temperature and pixel are defined properly when using uncalibrated images.
- If using Docker, verify that you have built the container correctly.
For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.
Future Development Plans
The developers of AiThermometer have exciting ideas for enhancements, including:
- Increasing the number of supported thermal cameras.
- Enhancing the people detection rate of OpenPose on thermal images.
- Improving robustness for measuring facial temperatures.
- Automatically detecting masks and other facial coverings.
Conclusion
At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.
Disclaimer
The information provided by AiThermometer is for informational purposes only. Please consult with a personal doctor before making any health or medical decisions based on this software.
License
This project is licensed under the terms of the MIT license and incorporates material from various projects, including OpenPose, OpenCV, and flirimageextractor.