Human seeker robot in hostile environments.
- What is this?
- Requirements
- How to use
- Controller panel
- Driving
- Speech to text
- Image processing
- Person detector
- License
- Contribution
- Authors
Luci is a human detector robot designed to locate people inside hostile enviroments such as gas, smoke or fire.
Features:
-
Easy to controll and operate.
-
Widely range of sensor and cameras.
-
Small and fast.
For running each sample code:
-
[C++]
-
Step 1:
cd code -
Step 2:
pip install -r requirements.txt -
Step 3:
python main.py -
Step 4: Isn't it beautiful?
Luci has a build-in controller panel web server. Her user-friendly interface allows the operator to see through Luci's eyes and sensors can drive where none can arrive.
Luci's has a wide range of sensors such as: gas sensor, distance, temperature, humidity, microphone and two cameras. You can see all the sensors displaying below:
In order to perform more accurate driving experience, Luci is controlled with a gamepad. This cool feature allows the operator to control each wheal movement and speed separately in order to get a better accuracy movement.
Luci's microphone can recognize voice and convert it to text. This feature is founded on the controller panel and allows the operator to read the live transcription of the room's voice.
The robot's eyes are formed of a wide-range camera and a thermal camera that combined using image processing provides an RGB realtime image with temperature coloring.
Finally, in order to drive and have a real time image from Luci, she has a build in video stabilization developed with software that uses a wide range of image stabilization such as: KeyPoint matching, Fourier image superposition and Optical Flow. Each age can be called by the operator in order to find the implementation that works best for each situation.
Luci's computer runs a Neuronal Network that detects people and animals in the room. This tool helps the firefighters to identify the number of people inside a building and combined with the thermal camera we can determinate the state of the subject.
MIT
Any contribution is welcome!!