Group Actions
Engagement
  • Author Author: dougw
  • Date Created: Date Created
  • Views 1471 views
  • Likes 10 likes
  • Comments 17 comments
Related
Recommended

ComputeCam and PicoPorch - Early Muddling

dougw

Intro

This is the second installment of Experimenting With Extreme Environments.

I have been struggling to navigate through a maze of long and twisty rabbit holes in the early stages of this project. Although some progress is being made, it still feels like there are a lot of unknowns to resolve.

This blog includes scattered tidbits of interest  about what I have been up to, just to show some progress, but most of the hard slogging and dead end efforts aren't exciting enough to make it into the blog.

Computers and Cameras

Here are some early descriptions of the computers and cameras that will be used in the project:

Camera Action

Here is a demo of 3 cameras in action:

The PicoPorch Panning System

Here is some preliminary work on the Pi Pico camera panning controller and Midas display:

Of course this needs to work with a Pico W, but the one I ordered has not arrived yet.

Discussion

There is still a lot of work to do to explain what was done to get these results, but there are a lot of aspects that are not finalized yet, so this blog has been more of a series of teasers about what is happening on the project and what is to come.

So far I have ordered a lot of parts, but there are still many that have not arrived. It is tough to design with parts that are not here and have no datasheets. I have designed and 3D printed at least 10 parts, but there are quite a few still not started.

One custom PCB has been received, the other one had to wait until a breadboard circuit was proven, so it is going to be late - hopefully it arrives before the project deadline.

Two Hammond boxes have been modified to house Raspberry Pi Zero 2W computers with Pi cameras. These boxes are handy since these modules get warm and 3D printed housings might sag with the heat. I have made some mods to one of the 3 clear lid Hammond enclosures, and started designing the insert structure, but there is a long way to go on that front.

The next immediate step is to design and build the motorized panning mechanism, even though the Pico W has not arrived.

Links:

    Parents

    • Thanks for sharing your progress and problems encountered. I'm not using the cameras, but I've checked the motionEyeOS image repository and it appears to be a dead project. There have been no updates for two years and it appears that an image for the compute module 4 (CM4) was never built.

      I would try to try the project https://github.com/motioneye-project/motioneye which seems more standard and I think it will work on the CM4 with the Raspberry OS. It's just a proposal, I haven't tried it. Good luck!

    • in reply to javagoza

      I will keep plugging away at the CM4 to see if I can figure something out. The standard Raspberry Pi OS has built-in camera support but I can't even get that working. You are right about the motionEyeOS - it doesn't even work with the Pi Zero 2, but I have found a 3rd party build that does work with Pi Zero 2.

    • in reply to dougw

      I was quite distant from the Raspberry Pi OS for two years or more and I have found that with the new Debian-based operating system Bookroom it has become a nightmare to install python-based applications. On the one hand it's great to have virtual environments for Python but many installable packages don't take it very well and there is a lot to do by hand to make them work. You always have the option of installing the previous Bullseye version.

    • in reply to dougw

      I used it with the Pi Zero 2 W a couple of years ago /members-area/personalblogs/b/ralph-yamamoto-s-blog/posts/raspberry-pi-zero-2-w---motioneyeos, but the fork I used also does not appear to have been updated since then.  I'll be interested to see if you get it working with the CM4.

    • in reply to javagoza

      With  Debian Bookworm any code that uses a library added via the pip utility should be added and executed inside a virtual environment.

      You can do the following:

      • Open a terminal window and create a new directory:
      mkdir MyWorkingDirectory
      • Enter the new directory with:
      cd MyWorkingDirectory
      • Create a new virtual environment that inherits all currently installed packages with:
      python -m venv --system-site-packages env
      • Enter the virtual environment with
      source env/bin/activate
      • Your command line prompt should now have a prefix of (env)
      (env) _
      • We can now use pip to install :
      (env) pip install ...
      (env) pip install ...
      ...
      • If you need to exit the virtual environment, execute deactivate from the command line.
      (env) deactivate
      • To reenter the virtual environment simply execute source env/bin/activate again from inside the  MyWorkingDirectory folder.
      source env/bin/activate
      (env) _


      I have had to do this to include the libraries that I am using in my solution, such as the TensorFlow library, to perform the classifications of the recorded audios.

    Comment
    • in reply to javagoza

      With  Debian Bookworm any code that uses a library added via the pip utility should be added and executed inside a virtual environment.

      You can do the following:

      • Open a terminal window and create a new directory:
      mkdir MyWorkingDirectory
      • Enter the new directory with:
      cd MyWorkingDirectory
      • Create a new virtual environment that inherits all currently installed packages with:
      python -m venv --system-site-packages env
      • Enter the virtual environment with
      source env/bin/activate
      • Your command line prompt should now have a prefix of (env)
      (env) _
      • We can now use pip to install :
      (env) pip install ...
      (env) pip install ...
      ...
      • If you need to exit the virtual environment, execute deactivate from the command line.
      (env) deactivate
      • To reenter the virtual environment simply execute source env/bin/activate again from inside the  MyWorkingDirectory folder.
      source env/bin/activate
      (env) _


      I have had to do this to include the libraries that I am using in my solution, such as the TensorFlow library, to perform the classifications of the recorded audios.

    Children
    No Data