From eLinux.org
Jump to: navigation, search

Please share your MinnowBoard-based projects here.

MinnowBoard Fish Picker-Upper

This is a robotics and computer vision demo by Scott Garman which uses the MinnowBoard and OpenCV to recognize and pick up an object.

Here is a video which describes how it operates.

Source Code

Yocto Project layer for this project on GitHub: meta-robotarm-opencv-demo

You will use this layer to build the robot-opencv-demo-image and install it to a microSD card which the MinnowBoard can boot.

Application sources on GitHub: minnowboard_fish_picker_upper

This python code is what controls the robot arm and runs the OpenCV object detection library to recognize and pick up the object.


The documentation for how to use this is based on the README files included in the above git repositories. If you're new to using the Yocto Project to build embedded Linux images, please run through the Yocto Project Quick Start guide to set up your system and test building your first image. Using layers is described in the Understanding and Creating Layers section of the Yocto Project Development Manual.

Once you've successfully built the robot-opencv-demo-image, copy the sources from the minnowboard_fish_picker_upper repository to your MinnowBoard's root filesystem and follow the documentation in the project's README file. You will either need to generate your own OpenCV object detection database or use the demo tarball which is linked to in that README.