An Underwater Motion Capture System That Does Not Cost $150,000

Version 1.0.0

Marvin’s summer internship project at the University of Washington’s Applied Physics Laboratory

Underwater tracking of ROV submersibles is difficult. On-board sensors are often inaccurate and GPS signal does not reach the depth that the ROVs operate at. Because of this, using external cameras to track and locate objects deep underwater is an effective method. However, since the only commercially available underwater camera system costs a LOT of money, this project is meant to be an affordable alternative.

Setting Up

Installation and environment setup for both the server and clients are explained here:

• Server
  • Installing Python3
  • Installing Python Packages
• Client Hardware
  • Out-of-the-box Pi
  • Attaching the Camera Module
  • Initial Configuration
• Client Software
  • Python3
  • Creating a Virtual Environment
  • Installing Packages

Running the Script

1. Run threading_server.py on the server device.

2. Run threading_client.py on each client device. Each client and the server should show a “connected” status message.

3. When prompted, type “image” on the server task window to request an image, or type “quit” to terminate the program

4. Received images will be in the four_way_images folder

5. Once images are received, user will be prompted once more for input, looping back to step 3

Additional Information:

This project is licensed under the BSD License:

• Licensing