← LEARN·PLATE 04·Tools

Photo overlap, the survey grade default.

Overlap is the photogrammetry parameter that quietly decides whether your map reconstructs cleanly or comes back with holes. Too little and the software can’t find tie points; too much and you’ve doubled your flight time for no benefit. The sweet spot is narrower than most pilots realise.

§ 01Try it

Drag the front and side overlap sliders below. Watch the photo footprints either crowd together (over-collection) or pull apart (gaps). The badge in the corner shows the quality assessment.

OPTIMAL

Good quality at sensible cost. Industry-standard for survey work.

Photos24
Capture window~29s
FIG.03Photo footprint grid. Each rectangle is one image; overlapping zones provide tie points for matching. Showing 75% front / 70% side overlap.
OVERLAP
The percentage of one photograph that is also visible in the next. Front overlap is along the flight line; side overlap is between adjacent flight lines.

§ 02Why overlap matters at all

Photogrammetry reconstructs 3D position by triangulating the same ground feature across multiple overlapping photographs. With 75% front and 70% side overlap, every interior point on your ground is photographed by at least 6–8 cameras from different angles — which is what gives the bundle adjustment enough constraints to converge accurately.

PC₁C₂C₃GROUND TRUTH · POINT P RECOVERED FROM 3 OBSERVATIONSFIG · TRIANGULATION FROM N CAMERAS

§ 03The recommended defaults

For most drone mapping work, the industry-standard starting points are:

Open terrain
70% front / 60% side — fastest, works for fields and farmland
General mapping
75% front / 70% side — the default in DroneDeploy, Pix4D, ODM
Urban / vertical structures
80% front / 75% side — needed for buildings, façades, infrastructure
Forestry / vegetation
85% front / 80% side — low-texture surfaces are hard to match
3D modelling / orbit
85% + oblique passes — need angles, not just nadir overhead

§ 04When overlap is too low

Below 65% front or 60% side, the demo above flips to RISKY. Here is what actually breaks:

Coverage gaps
Areas the cameras never overlapped enough — these become holes in the orthomosaic.
Failed bundle adjustment
SfM cannot find enough common features to lock down camera poses.
Bad heights
Even when reconstruction succeeds, vertical accuracy is the first thing to degrade.
Texture-poor surfaces
Snow, water, gravel, freshly-tilled fields — these need extra overlap to compensate.

§ 05When overlap is too high

Going from 75/70 to 85/80 doubles photo count and flight time, but only marginally improves reconstruction quality on normal terrain. The exception is anything with little natural texture (snow, sand, water) where the extra redundancy buys you a workable model.

§ 06Tie points — the math underneath

Each overlapping pair of photographs produces hundreds to thousands of tie points: features (corners, textures, edges) that appear in both images. The photogrammetry engine — usually OpenSfM, Pix4D’s engine, or COLMAP — uses these to triangulate the camera poses and the 3D positions of every visible feature.

Below ~65% overlap, the number of tie points per image pair drops sharply, and so does the geometric strength of the reconstruction. That cliff is what the badge on the demo is warning you about.

§ NEXTRelated field-guide pages

Map your first site in the next hour.

Free to start. No credit card. No sales call. The full mapplot mission planner runs in your browser.

Start free