• PeriodicallyPedantic@lemmy.ca
    link
    fedilink
    arrow-up
    0
    ·
    14 days ago

    Idk, I think they can probably do a reasonable job tracing a drone back to it’s takeoff location, and then tracking the person who brought it there back to their home, with decent coverage.

    That said I don’t know if I’m overestimating their ability regarding machine learning and AI - this is probably fairly labour intensive unless they’ve don’t a good job preparing all their data and they have plenty of compute.

      • PeriodicallyPedantic@lemmy.ca
        link
        fedilink
        arrow-up
        0
        ·
        14 days ago

        I mean if they can track a car around the city using their cameras, they can probably do the same for a drone. Even if you fly up out of view they could look for drones in a radius flying down into view and connect the dots.

        • einfach_orangensaft@sh.itjust.works
          link
          fedilink
          arrow-up
          0
          ·
          14 days ago

          Resolution of those cameras is limited and they are usually angled down. Also they are not on private property. So it can be assumed that max detection rate of a average sized done is maybe 80m tops, after that the drone itself is smaller than a singular pixel of the cam. Meanwhile range of those drones can be multiple km.

          If u had said they can use the integrated wide band SDR recordings of cell towers to triangulate the remote controller…i would have given u a point, but triangulating a drone using cameras that look down is not credible.