Autonomous Surface Vehicle in Search and Rescue Process of Marine Casualty using Computer Vision Based Victims Detection

Code Implementation used in my research refer to my IEEE paper

if you don't have access to IEEE there is copy of my paper on my linkedIn Profile

---

This is My Final Project (research) for my Bachelor Degree. All code implemented using C++, except the Computer Vision part is using python.

This Project is consist of some ROS node:

• Control System
  • Mission Planning and Track Target using Computer Vision
  • Obstacle Avoidance using Ultrasonic Sensor
  • Return To Home
• Computer Vision
  • Using YOLOv4 to detect human on the water (Marine Casualty Victims)
  • Model converted to tensorRT so can running on mobile laptop GPU
• Localization
  • Global Localization using GPS
  • IMU, Compass, and Gyro for orientation
  • Combined data implemented to TF
• Ground Control and Monitoring
  • Implemented using QT and ROS RViZ
• UDP Communcation to Microcontroller (STM32F4)
  • Drive Motor
  • Drive LED Indicator
  • Receive any radio communication from RC Controller
  • Emergency Stop