Barbaros USV obstacle detection and avoidance algorithm

This algorithm is an algorithm that detects obstacles via yolov5 in line with the barbaros usv system autonomous obstacle avoidance algorithm. It gives the following outputs according to the position of the obstacle over the safe zone it has drawn.

Run it on your computer

download the project

  https://github.com/Whitte10/BarbarosAIsystem

go to project path

  cd my-project

install required packages

  pip install /r requirements.txt

run the code

  python3 detect.py 

Note:if you are in ubuntu specify main folder location on line 8

Specifications

• Ubuntu support (change on line 8 required)
• Windows support
• Compatibility with different inputs (by inserting images into testimages)
• Compatibility with different resolutions

Associated Projects

Here are some related projects

YoloV5

Technologies Used

İslemci: Intel(R) Core(TM) i7-10750H CPU @ 2.60GHz 2.59 GHz

GPU: Nvidia RTX2060

Lessons Learned

In the test images, the horizon line and the ship lines are sometimes confused because the POV images also include parts of the ship where the camera is mounted. In order to solve this error, the edges detected at the bottom of the picture were not included in the calculation and the horizon line was tried to be calculated in this way. And our model need more training because sometimes it can not detect obstacles accordingly

Screenshot

Ros detection

Real World Dtection