MJPG-streamer VS BirdNET-Pi

2. https://github.com/jacksonliam/mjpg-streamer/

mjpg-streamer uses MMAL to access the Raspberry Pi Camera data, but that's not implemented on ARM 64-bit Linux.

cd ~ sudo apt-get install cmake libjpeg62-turbo-dev:armhf libjpeg62-turbo-dev imagemagick ffmpeg libv4l-dev cmake git clone https://github.com/jacksonliam/mjpg-streamer.git cd mjpg-streamer/mjpg-streamer-experimental export LD_LIBRARY_PATH=. make sudo make install

I think I was able to solve my own issue. I'm now using https://github.com/jacksonliam/mjpg-streamer to start a stream. It creates a minimal web interface at http://pi-address:8080/, and the path /?action=stream can be accessed from MotionEye.

I need to stream all three of them to a Windows computer, so the maximum bandwidth is about 100mbit. I have tried "motion", but the performance is abysmal. I get not more that one frame per second and the latency is 3s. I used https://github.com/jacksonliam/mjpg-streamer with one camera. Its super fluid with a latency of maybe 200ms, and full 30fps. But the stream takes 60mbit/s bandwidth so it is impossible to use all three cameras. I tried changing the jpg quality but mjpg-streamer crashes.

Thanks for your interest! I will opensource this project soon, but in case you want to start off now, here's some resources I used: - MJPG-Streamer https://github.com/jacksonliam/mjpg-streamer - .NET 6, targeting Linux (tutorial:) - dotnet package on linux - My custom designed case (beta!): Bottom part - Top part

Have you looked into mjpg-streamer (or other forks, this was the first Google result)

Yup! There's an open source software called BirdNet, and someone packaged it up to make it easy to run on Pis - BirdNet Pi.

I had https://github.com/mcguirepr89/BirdNET-Pi running last summer with great success. Might be worth a look.

BirdNET-Pi

The birders are super ahead of the curve. Check out this Rpi implementation of BirdNet: https://github.com/mcguirepr89/BirdNET-Pi

Some other options for low-cost backyard sound/biodiversity monitoring are:

https://www.openacousticdevices.info/

If you do want to use a raspberry pi and the mics you build, check out this project: https://github.com/mcguirepr89/BirdNET-Pi/

I have I've used rpis for multiple things, but by far the most interesting thing I've done with them is as a [birdsong classifier](https://github.com/mcguirepr89/BirdNET-Pi). That requires an external microphone _of some sort_, and is prone to certain pitfalls (e.g. all sorts of background noises trigger false positives on bittern).