
Smart Controlled Unmanned Water Vehicle for Human Detection
Software as a Service
NoCategory
Thesis DesignClients
Mapua University
Techstack
PythonYOLOv4OpenCVRaspberry PiFlask
Purpose
Develop an unmanned water vehicle that automatically detects people in aquatic environments and reports their location, supporting search-and-rescue and water-safety scenarios where sending a human operator is slow or unsafe.
Description
A Raspberry Pi on board the vehicle runs YOLOv4 object detection against a live camera feed, with capture and inference split across threads to keep frame rates usable on embedded hardware. When a person is detected, a GPS module and SIM800L cellular modem send location alerts by SMS and email, while a Flask web interface streams the annotated video and logs each detection with processing times to CSV.




