[pjsg]

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.