Drone photogrammetry software runs the same reconstruction pipeline as any other photogrammetry tool: feature matching, camera solving, dense reconstruction, and the outputs that follow. That shared mechanism is covered in What Is Photogrammetry Software?; this piece stays with what changes when the photographs come from a moving aircraft instead of a tripod or a handheld camera. For a commercial pilot, most of what goes wrong on a mapping job traces back to a decision made before the software ever opened a file.
Overlap is a flight-planning decision, not a processing setting
Front and side overlap between adjacent photos determine whether the software will have enough shared coverage to solve the geometry at all — and no setting inside the processing software can invent overlap that was never captured. A grid pattern flown at a fixed altitude works fine over flat ground; the same grid over a site with real elevation change quietly drifts the ground sample distance and the effective overlap as the aircraft gets closer to or farther from the terrain. A corridor job, like a stretch of pipeline or a transmission line rising and falling with the land, usually needs terrain-following flight rather than a single constant altitude, or the overlap that looked fine on a laptop during planning turns out to be thin exactly where the terrain climbed.
Ground sample distance decides what the client can see later, not what they saw on the screen
Ground sample distance — the real-world size each pixel represents — is set by altitude, sensor size, and focal length before the aircraft ever takes off, and it cannot be recovered afterward by zooming into the finished ortho. A request like "I need to see the cracking on that panel" is a GSD requirement, not a processing request, and it has to be matched to a flight altitude and lens combination before the mission, not discovered as a disappointment during review. Flying lower for finer detail also means more photos, more overlap to maintain, and more battery cycles to cover the same ground — a tradeoff that belongs in the quote, not just the flight log.
RTK, PPK, or ground control: three ways to hand the software a position
Real-time kinematic corrections tag each photo with a position as it is taken, but only as reliably as the live correction link holds up — a link that can drop near structures, tree lines, or terrain that blocks a radio or cellular signal. Post-processed kinematic logging records raw satellite data onboard and corrects it after the flight against a base station or reference network, which sidesteps the live-link problem but means the corrected positions are not available until after landing. Ground control points are independent of either GNSS method: physical, surveyed marks placed and measured before the flight, then identified in the photos afterward. A pilot relying on RTK or PPK alone is trusting the aircraft's own positioning chain end to end; a job that needs an outside check still benefits from at least one control point that never passed through the aircraft's GNSS receiver at all.
Rolling shutter and lens calibration: artifacts a tripod never produces
Many mapping cameras use a rolling shutter, which exposes an image line by line rather than all at once. On a stationary subject that is invisible; on an aircraft banking through a turn, correcting for a gust, or vibrating against its own frame, it can skew straight lines and distort geometry in ways a stationary tripod shot never will. Locking exposure and confirming focus at the start of a mission matters more in the air than on the ground, because a lens that drifts slightly with vibration or temperature across a long multi-battery flight introduces a calibration mismatch between the first photos of the day and the last — exactly the kind of inconsistency that shows up later as a seam or a soft patch in the ortho.
Battery cycles turn one flight into several missions
Anything beyond a small site gets split across battery packs, and each swap is a seam risk, not just a pause. Light changes while a battery is off charging — clouds move in, the sun drops another few degrees — so photos from the last battery of the day can carry a visibly different exposure than the first, even over the same ground. Resuming a grid line by eye, or trusting the flight app's return-to-last-line feature, can leave a sliver of missed overlap that is easy to miss on a small planning screen and expensive to discover after landing. Treating each battery as its own labeled flight segment, and checking coverage continuity at every swap rather than only at the end of the day, catches most of this before it becomes a processing problem.
Why drone jobs fail in different ways than a ground-based capture
A terrestrial photogrammetry session usually covers a small area in a short window, which limits how much can move between photos. A drone mapping mission covers a wide area over many minutes, which gives wind, wildlife, vehicles, and people far more opportunity to shift position between one overlapping pass and the next. Wind introduces roll and pitch that a gimbal only partially corrects, feeding the reconstruction photos that were not quite where the flight log says they were. Prop wash can visibly disturb a water surface directly under the aircraft during a survey of a pond or a tailings pond. Flying near steel structures or heavy equipment can introduce magnetic interference that throws off the heading used to seed the initial camera pose, which the software then has to fight through during alignment rather than accept as a clean starting guess. None of these are photogrammetry problems in the software sense — they are aviation-and-environment problems that show up as photogrammetry symptoms.
What a commercial handoff needs that a casual flight doesn't
A client paying for a deliverable is owed more than the files. Keep the flight log, the camera and lens settings actually used, any RTK or PPK correction files, and a note on which battery or flight segment covered which part of the site — not because a regulation demands it, but because that record is what lets you explain a defect six months later instead of guessing. State plainly what was checked and what was not; a client who knows a corner of the site had thin overlap can plan around it, while one who finds out from a bad measurement later will not trust the next map. The aircraft is what makes this job different from a tripod job, and the paperwork is what makes it a professional one.
If a flight-planning tradeoff on a specific site is still unclear, ask the follow-up on Ask Mav at pilotledger.com.