UAV Margin Calculator

1 · Aircraft & sensor

Platform / payload

Sensor width (mm)

Sensor height (mm)

Image width (px)

Image height (px)

Focal length (mm)

Min. photo interval (s)

Camera orientation

2 · Flight parameters

Altitude AGL (m)

GSD (cm/px)

Frontlap (%)

Sidelap (%)

Speed (m/s)

Shutter speed 1/x (s)

How to set the shutter speed

Why it matters. While the shutter is open the aircraft keeps moving, so every point is smeared over speed × exposure time. Divided by the GSD, that gives the motion blur in pixels:

blur (px) = speed (m/s) × (1 / shutter) / GSD (m)

Targets: ≤ 0.5 px is sharp (best for 3D and accurate tie points); 0.5–1 px is acceptable for orthomosaics; > 1 px loses detail and accuracy. Enter here the shutter speed you will use (or read ExposureTime from the EXIF of a test photo); the line above tells you the slowest shutter that stays sharp.

Setting it in the app (GS RTK, GS Pro, Pilot 2):

Use shutter priority (S / Tv) or manual exposure, and enter at least the value shown above. Sunny: 1/1000 – 1/1600 s is typical for a 1" camera at 8 – 10 m/s.
Let ISO rise (auto ISO, capped at about 400 – 800 on 1" sensors) rather than slowing the shutter — a little noise hurts less than blur.
Aperture f/4 – f/5.6 is sharp on most mapping lenses; on the P1 the fastest shutter (1/2000 s) needs f/5.6 or smaller.
Overcast or low sun: fly slower instead of using a slower shutter. Blur halves when speed halves.
Prefer cameras with a mechanical or global shutter. Electronic rolling shutters add extra distortion at speed.
Multispectral cameras (P4 Multispectral, Mavic 3M, MicaSense) expose automatically and do not let you set the shutter. Use this field to check blur for the exposure they actually used (EXIF), and keep speed low in dim light.

3 · Product & edge target

Deliverable

Min. frontlap at edge (%)

Min. sidelap at edge (%)

How to set the edge targets

What they mean. Points inside your area get the frontlap/sidelap you fly. Points on the boundary only get photos from the inside unless the route continues past the boundary — that extension is the margin. The edge targets are the lowest frontlap and sidelap you accept on the boundary of the area you must deliver. The calculator finds the margin that reaches them.

Overlap ↔ number of photos. Overlap o along one axis means a point is seen by about 1/(1 − o) photos on that axis. Images per point = (photos along) × (photos across).

Overlap	50 %	60 %	67 %	70 %	75 %	80 %	85 %	90 %
Photos / axis	2	2.5	3	3.3	4	5	6.7	10

Suggested values (edge frontlap / sidelap):

Deliverable	Edge target	Why
Orthomosaic, flat / built-up	70 / 60	≈ 8 images per edge point: good tie points and seamlines. 60 / 50 (5 images) is the minimum.
Vegetation, crops, multispectral	75 / 65	Repetitive texture and wind make matching hard; reflectance maps need many views.
3D model / DSM	= flight overlap (e.g. 80 / 70), on object tops	The model edge should be as good as the centre. Outward façades also need camera stations outside the boundary.
LiDAR	dwell 100 %, sidelap = flight sidelap	Avoids using only the noisy outer edge of one strip at the boundary.

Rules. (1) The edge target cannot exceed the overlap you fly — raise the flight overlap first. (2) Keep edge images per point ≥ 5 (e.g. 60 / 50). (3) A higher target means a larger margin, more lines and more flight time. (4) A margin in metres grows with altitude, because footprints grow. (5) Drawing the polygon larger than the deliverable boundary has the same effect as a margin.

Tallest object near boundary (m) — trees, roofs

Footprints shrink on top of tall objects, so overlap on rooftops/canopy is lower than on the ground.

4 · Survey area (for coverage map)

Width is measured perpendicular to the flight lines (the direction the aircraft steps from line to line). Length is measured parallel to the flight lines. In DJI apps the line direction is set by the course angle (main course). For an irregular polygon, use the size of its bounding box after rotating it to the course angle.

Width across flight lines (m)

Length along flight lines (m)

Evaluate coverage on

Verification — does every point of the area get enough overlap?

Checked by simulating every photo of the plan with the margin you set (see the coverage map) and counting the photos that see each point of the survey area. Assumes nadir photos, a constant height above the evaluated surface (terrain follow, or set the altitude above the highest ground), and no wind drift. Over hilly terrain without terrain follow, overlap drops where the ground rises.

Flight geometry

Recommended margin

DJI apps apply one margin value as an outward buffer around the whole polygon, so the recommendation is the larger of the along-track and across-track requirements, rounded up to a whole metre.

Your margin → edge overlap

Margin (m)

Margin (m) — enter the same number in the DJI app.

Margin vs. effective overlap at the boundary

Coverage map — how much overlap does each spot of the survey area get?

Colour the map by

Survey area (your deliverable) Area + margin (where the drone flies) │ flight lines · • photo positions · flight direction ↕ Outside the survey area is dimmed

Survey area by overlap class

Hover or tap the map to read the overlap at any spot. The app may start the first line anywhere within one line spacing of the margin edge. Six placements are tested and the map draws the worst one, so what you see is what is guaranteed.

Field summary

How the calculation works

Footprint and spacing (camera):

GSD = H × pixel pitch / focal length Footprint across W = GSD × pixels across track Footprint along L = GSD × pixels along track Photo interval B = L × (1 − frontlap) Line spacing S = W × (1 − sidelap) Images per point (interior) ≈ (L/B) × (W/S)

A point in the interior is seen by about L/B photos along track — but photos are whole numbers, so some points get ⌊L/B⌋ and others one more. A flight at overlap o therefore guarantees m = ⌊1/(1 − o)⌋ photos per axis at every point (80 % → 5, 75 % → 4, 70 % → 3, 60 % → 2). The calculator works with these guaranteed counts.

A point on the boundary only gets photos from one side unless the route continues past it. Every station within L/2 of the point sees it, so with margin M the stations that can see it lie in an interval of length min(M, L/2) + L/2. Whatever position the app picks for the first photo/line, that interval holds at least

n_edge(M) = ⌊ ( min(M, L/2) + L/2 ) / B ⌋ photos along track (same with W and S for lines across track)

With M = 0 the boundary gets only about half the interior count. With M = L/2 (half a footprint) it is as good as the interior. Solving n_edge ≥ m for the edge target o:

m = ⌊ 1/(1 − o) ⌋ M = m × B − L/2 (along track) M = m × S − W/2 (across track) Recommended margin = max of the two, rounded up to whole metres

The verification panel checks this independently: it places every photo and line of the plan for six different alignments of the first line and counts, at 1,000 sample points along each axis of the area (always including the boundary), how many photos and lines cover each one.

For objects of height h, footprints on their tops scale by k = (H − h)/H, so W and L above are replaced by k·W and k·L. This is why tall buildings and tree canopy at the edge need extra margin, and why overlap on rooftops is lower than the value set in the app. A half-footprint margin also puts camera stations outside the boundary so façades facing outward are actually imaged.

The coverage map draws the worst of these alignments. Real apps may place lines and triggers slightly differently, add turn segments, or clip the margin, so treat the map as a model and always check the planned route in the app.

Rule-of-thumb image counts: ≥ 5 images per point for reliable matching; 3D/forest projects usually need ≥ 9–15 (e.g. 80/70 or 85/75). Preset camera values are taken from published specifications — check them against your own EXIF data if you need high precision, or use “Custom”.

UAV Flight Plan Margin Calculator

Help centre

Verification — does every point of the area get enough overlap?

Flight geometry

Recommended margin

Your margin → edge overlap

Margin vs. effective overlap at the boundary

Coverage map — how much overlap does each spot of the survey area get?

Survey area by overlap class

Field summary