Unclear what the relationship is with the Global Meteor Network (https://globalmeteornetwork.org/) that has people build meteor cameras, often based on RPis, and then contribute the data so that meteor orbits can be determined.
This site: https://tammojan.github.io/meteormap/ shows the meteors detected over the last 24 hours. You can see that the UK is pretty well covered with cameras as are some parts of Europe. The US is rather sparser -- with only Arizona having good coverage.
Building a camera is fairly easy and is under $200 -- most of the parts can be ordered on Aliexpress.
Even cheaper are esp32 boards. Seeed has tiny esp32c3 with camera included for $14 USD. Granted the camera module isn't the greatest but it may be good enough for sky tracking, there are also other camera modules.
I feel like RPis are overkill unless the data is being processed in situ (even then the rpi may be overkill compute) beyond just packaging it up digestibly to be crunched elsewhere.
There are two challenges -- one is getting a sensitive enough camera to be able to record meteors. The Sony 'starlight' sensors have the required sensitivity. The next is to handle the data stream from the camera and to extra the meteor tracks from the video stream. Accurate timing of each frame is also important so that the meteors can be tracked from multiple cameras and the orbits calculated.
At my location, on most nights, I have far more aircraft crossing the field of view than meteors -- and that needs a far amount of compute to extract the signal from the noise.
A typical set of output from a nights run: https://globalmeteornetwork.org/weblog/US/US001N/US001N_2024... -- the first images is the all the meteors from the night (but it is clear that there are some aircraft there). The second image is all the tracks seen overnight and you can see that it is almost entirely aircraft.
I feel this way about most uses of RPis, but if it makes it more accessible and easier to use without the limitations of esp32 type boards then at least people are playing with things.
However, from TFA: "The Raspberry Pi computer analyses the video feed in real time to search for moving objects, "
I see the accessibility angle though I suppose if someone made a turn-key esp32-based sky tracker that might help get them adopted, save people money, and free up their pi.
How I should read the metor lines? What end of the line is impact point? It is really an impact point or last point where the light was observered. How I can search for a metheor using this map?
The meteors go in the direction from darker to white. So the white part of the trail is the lowest and last part of the trail.
Nearly all meteors do not actually impact the ground -- they burn up in the atmosphere. However, sometimes you get the end point quite low in the atmosphere -- then you can try and predict roughly where it will fall.
I wonder if SpaceX could put cameras on the far side of the Starlink satellites to make a network gazing outwards. To help compensate for the frustration they are causing astronomers…
„As of early March 2024, it consists of over 6,000 mass-produced small satellites in low Earth orbit (LEO)[5] that communicate with designated ground transceivers. Nearly 12,000 satellites are planned to be deployed, with a possible later extension to 34,400.“
The whole thing could be considered a 360 degree camera. I have no idea how good the theoretically achievable resolution would be overall.
I wonder how long the sensors on those cameras can survive being pointed up at the sky like that. Surely the image quality would deteriorate over time as the sun destroys the sensor until it eventually stops working.
I'd be interested in how they protect the sensors during day. After briefly skimming the sources it doesn't seems to be present any code to activate/deactivate an external shutter or servo to rotate the camera away from sunlight. There is code to turn the camera on and off but it doesn't seem to do anything else beyond that.
Fully powering off the camera will help, but I imagine with direct sunlight going through the lens and hitting the sensor the heat could still damage it, despite it not recording.
I have a sky camera (use it for weather and sky observations; my airspace is a bit busy for easy meteor observations). So far my sensor seems to be fine despite an unblinking gaze upwards.
Looked through the Github files and I don't see anything that describes the hardware. This web page shows a picture of the insides of the camera housing (click "Next" once to see the picture):
Probably the "Siemens 1/2" F1.2 4-12mm Varifocal Auto Iris C Mount CCTV Lens" since the AutoIris port on the camera is connected to the lense in the picture.
The video capture is an Easier CAP USB capture device, like this:
The Pi is a Raspberry Pi 4 Model B according to the text but the photo caption says it's a Pi Mk 3, maybe either will work. It has a Pi Hat adapter board on the top (POE plugs into it and it seems to have a relay that connects to the camera, maybe to control power to the camera). The camera outdoor housing is made by Genie, looks like a TPH-2000, no longer made, runs on 240V and has a heater:
Most satellites and debris don't reflect that much light, plus there's already pretty extensive tracking of any object up there above a certain size by the USSF (formerly USAF) spacefence.
This site: https://tammojan.github.io/meteormap/ shows the meteors detected over the last 24 hours. You can see that the UK is pretty well covered with cameras as are some parts of Europe. The US is rather sparser -- with only Arizona having good coverage.
Building a camera is fairly easy and is under $200 -- most of the parts can be ordered on Aliexpress.