fish_vision

We are building the computer vision system for a very fast robotic fish. Target to avoid obstacles with vision.

Hardware

• Raspberry Pi 4 (Model B 8GB RAM)
• Raspberry Pi Camera V2

Software

• Ubuntu 16.04 + ROS Kinetic, image can be downloaded from Ubiquity Robotics.
• Python 2.7.12
• OpenCV 3.3.1
• Numpy 1.11.3

How To Run

Tools

• In "tools" folder, "video_capture.py" can be run in Raspberry Pi for capturing videos continuously.

Playground

"playground" folder contains code that has no hardware or OS dependency. Once set up environment with Python 2.7.X, OpenCV 3.3.1, Numpy 1.11.3, you shall be able to run them.

ROS nodes

1. roslaunch raspicam_node camerav2_1280x720.launch in Raspberry Pi.
2. rosrun obstacle_detector_node obstacle_detector.py in Raspberry Pi.
3. rosrun obstacle_avoidance_node obstacle_avoidance.py in Raspberry Pi.
4. rosrun debug_node vision_debugger.py in another computer.