That is true for what we nicknamed a "suicide camera", ie one straight in front of the vehicle. This is because loses a point of reference, when the vehicle arrives at the camera it is infinitely wide, the same as every vehicle, it cancels out a dimension.I have been doing some perspective drawing exercises and I have a feeling that even with 3 photos, you can only determine that the car is approaching at a constant speed, but you can't determine its absolute distance and speed. You will always have the unknown from "how big is this car?".
In my concept, the increase in size is exponential through time... is your equation also exponential?
You would also need a straight enough road, to trust that distance calculation with visual "foreshortening" only.
At least you can know how much faster one individual car is going, compared to the average.
Photographing from above or from the side adds that dimension back in but loses another unless you have a known height above the road or sideways offset. However from three photographs with a sideways offset you are adding another dimension in because you are tracking two points on the vehicle which have separate curves. So compared to a suicide camera you lose one dimension but gain two.
...... or something like that, it was a long time ago,
From three 2D photos you can't tell if a vehicle is at constant speed without more data. The assumptions we had to make was constant speed, straight line travel in all dimensions and to a lesser degree that the vehicle was horizontal (this is less important if the height of the camera was similar to the height of the measured points on the vehicle).
Statistics: Posted by pidd — Sat Feb 21, 2026 2:47 am