Tree


README.mdcommits | blame
README.txtcommits | blame
coco.namescommits | blame
frozen_inference_graph.pbcommits | blame
main.pycommits | blame
on_dataset.pycommits | blame
output.avicommits | blame
recorded vedio.pycommits | blame
single_img.pycommits | blame
sms.pycommits | blame
ssd_mobilenet_v3_large_coco_2020_01_14.pbtxtcommits | blame
v1.mp4commits | blame
yolov3.cfgcommits | blame

README.txt

Human-Detection-Using-Drone-During-Disasters

This project focuses on detecting humans in disaster scenarios using three
different methods. The system is designed to send a message to a predefined
number whenever a human is detected. The detection methods include:

Live Camera Feed: Utilizes the laptops live camera to detect humans in
real-time. AI-Generated Dataset: Employs a custom dataset simulating people
trapped in earthquake scenarios to train AI models for human detection.
Drone Footage: Analyzes video recorded by a DJI Naza M Lite drone to
identify humans in the footage. We used Vonage for sending messages and
Visual Studio for project development. The objective of this project is to
enhance disaster response times by promptly alerting rescuers to the
presence of humans in disaster-stricken areas.

In each scenario, the software sends an alert message to a predefined
number upon detecting a human, ensuring a faster response in disaster
situations.