Image Ground Footprint

Also known as image footprint · ground coverage per photo · frame footprint · photo coverage area · how much ground does one drone photo cover · acres per image · sensor footprint on the ground · ground coverage width · footprint from GSD

A=GSD2nxnyA = \mathrm{GSD}^{2} \, n_x \, n_y

Enter your known values, leave one input blank, and solves for the missing one. Try different units for next level excitement!

Learning zone

Once you know what one pixel covers, the frame follows by multiplication: the ground width along an edge is the GSD times the number of pixels on that edge, and the area is the product of the two. Nothing subtle happens here, and that is exactly why it is worth being careful — this is where a flight plan is either right or quietly broken.

Which edge is which decides the whole plan. A frame has two ground dimensions and they do different jobs. The edge lying ALONG the flight direction is what the endlap eats, and what is left of it is the air base — the distance flown between shutter releases. The edge lying ACROSS the flight direction is what the sidelap eats, and what is left is the flight line spacing. Get them the wrong way round and the flight either wastes half its battery flying lines far too close together, or leaves gaps that no amount of processing will fill. Most mapping missions fly with the LONG sensor edge across track, because a wider swath means fewer lines, fewer turns, and less time spent at the ends of lines doing nothing useful. A few payloads and a few planners do the opposite, so check rather than assume.

The footprint is not the coverage. Neighbouring frames overlap heavily by design, so the ground that is genuinely NEW in each exposure is only the endlap-and-sidelap remainder. At 75% endlap and 65% sidelap that is 0.25×0.35=8.75%0.25 \times 0.35 = 8.75\% of the frame — under a tenth. That is the number to use when estimating how many frames an area needs, and it is why the image count for a site is so much larger than a naive area-divided-by-footprint calculation suggests.

Two geometric assumptions sit under the multiplication, and neither is exactly true in the field. The first is that the camera is looking straight down. A nadir frame lands as a rectangle; tilt the camera even a few degrees and the footprint becomes a trapezoid, longer on the far side and with a GSD that varies across it. The second is that the ground is flat. Over relief the near side of a frame is closer to the camera than the far side, so the footprint is neither rectangular nor uniformly sampled. Flight planners handle this by working at a mean terrain elevation and adding margin, which is fine until the terrain is steep, at which point terrain-following becomes the honest answer.

One good habit that this page supports: quote the footprint in acres or hectares when you talk to a landowner. "Each photo covers about 1.4 acres" lands in a way that "3774 square metres per frame" does not, and it makes the image count for their field immediately believable. It also makes the cost of a finer GSD visible before anyone has committed to it, since halving the GSD quarters this area.

Image Ground Footprint
A=GSD2nxnyA = \mathrm{GSD}^{2} \, n_x \, n_y
nxnyGSDA
Where
  • AA= Ground area covered by one frame ()
  • GSD\mathrm{GSD}= Ground sample distance (per pixel) (cm)
  • nxn_x= Pixels across the sensor (one edge)
  • nyn_y= Pixels down the sensor (other edge)
Missing one of these? Work it out first, then come back