Videos












Technologies

Nvidia Jetson Nano

Image by: Build5Nines

NVIDIA® Jetson Nano™ Developer Kit is a small, powerful computer that lets you run multiple neural networks in parallel for applications like image classification, object detection, segmentation, and speech processing. All in an easy-to-use platform that runs in as little as 5 watts.

OpenCV

Image by: OpenCV

OpenCV is a library of programming functions that allowed the Fire Scout Team to process images and videos. These images and videos captured through Opencv would then be ran on our AI models.

Visual Studio Code

Image by: VSCode

Visual Studio Code was the IDE choosen by the Fire Scout Team to develop our code and to also intergrate with our GitHub repo. Visual Studio Code allowed our team to compare any changes on our current indevelopment branch along side with the master branch in our repo. This IDE also provided many other tools to detect any mistakes we may have made.



Python

Image by: juniortech

The python programming language is the only language we used to develop our project. Python was choosen based on its easy to implement infastructure and python makes it easier for future developers to understand why and what was coded.

Git + GitHub

Image by: freeCodeCamp

Fire Scout choose git as a version control system that allows the team to manage and keep track of source code. GitHub is a cloud-based hosting service that helps us to manage our gti repository.

TensorFlow

Image by: Tensorflow

TensorFlow is a free and open-source software library for machine learning. It can be used across a range of tasks but has a particular focus on training and inference of deep neural networks.



Requirements



Classification

If an image has a fire contained within it, then the image classification model classifies it as a fire. Otherwise it classifies the image as no fire.

Object Detection

Object Detection surrounds a desired object with a bounding box, along with a label and a prediction score associated with that object.



Image Segmanetation

Image segmentation breaks down a photo at the pixel level and seperates the fire from the rest of the image by assigning a different pixel color to the fire.



Phantom 3 Pro

Image by: DJI

Matrice 200

Image by: DJI




Path Planning

Path plannig helps to track the path of the drone and the fire by placing different collored pins for each. Blue is for the drone and the red circles are fires that have been detected.

SDR

The Software Defined Radio is a radio communication system that helps to transfer data from the drone to the ground station. This is espically useful in areas that have no internet connection or systems that can not utilize wifi.

Thermal Images

Red Hot (Fusion)

White Hot


If the drone finds itself in a situation or enviorment where a regular HD camera is not useful, it can switch to the FLIR thermal camera. This camera has the option to switch between three modes, Red Hot(Fusion), Green Hot, or White Hot.

Green Hot