My cat Tuco co-stars on my electronics and reverse engineering streams, but really, he deserves his own show. This project is a robotic camera for my cat, which will stream on Twitch with interactive control. The robot consists of a moving "flyer" portion with a camera, gimbal, and sensors. Four "winch bots" hoist it around the room on cords, from the top corners.Code and designs at:https://github.com/TucoFlyer
Just barely reached a pretty nice milestone today! With the computer vision object detection and tracking code mostly working, and the new Rust-based gimbal code coming together, I can just barely start to close the loop and start aiming the gimbal based on data from the vision subsystem. Here's some raw footage from tonight's experiments:
Finally after a few wrong turns and a whole lot of CAD work, I have a happy little winch that should be able to (with a few friends) move our camera bot around in style!
I've been a bit light on putting together documentation for this project so far, but the work continues and you can follow along with it on YouTube and Twitch. The archive from our latest stream is embedded below, where we hack on the motor control loop a bit. It was a fun stream with plenty of winching and plenty of catting.
Happy hacking folks, more detailed updates on the winch soon now that the first one is largely done, and I'll be continuing to stream the work in progress which you can get real-time notifications about via the @scanlimelive account.
If anyone's curious to see the project's progress in real-time (or skip around, naturally) here's the archive from the latest live stream where we get the winch going using the new motor drivers and power supply.
- Picavet rigging - https://publiclab.org/wiki/picavet-rigging - low tech traditional counterweights for camera stabilization - Have you considered the vision processing unit - Intel® Movidius™ Myriad™ X VPU to streamline your system? Vision bonnet for the AIY-Vision system is available! Myriad X is the industry leader for on-device deep neural networks and computer vision https://www.raspberrypi.org/magpi/aiy-projects-vision-kit/
Considerations for simplification ---
- Picavet rigging - https://publiclab.org/wiki/picavet-rigging
- low tech traditional counterweights for camera stabilization
- Have you considered the vision processing unit - Intel® Movidius™ Myriad™ X VPU to streamline your system?
Vision bonnet for the AIY-Vision system is available! Myriad X is the industry leader for on-device deep neural networks and computer vision https://www.raspberrypi.org/magpi/aiy-projects-vision-kit/
JimInAshland