GSD Calculator
Ground sample distance is the ruler printed invisibly on every aerial photograph: how many centimetres of Earth each pixel represents. Enter your flying altitude, lens and sensor, and this calculator returns the GSD — and the width of ground each photograph covers, the two numbers every mapping quote should state before it states a price.
Similar Triangles at Altitude, Priced
The geometry is the camera's own, scaled: the triangle from lens to sensor (focal length tall, sensor-width wide) is similar to the triangle from lens to ground (altitude tall, footprint wide), so the footprint-to-altitude ratio equals the sensor-to-focal ratio, and dividing the footprint among the image's pixels gives GSD = (sensor width × altitude) ÷ (focal length × image width), in consistent units. Every term is a lever a mission planner actually holds: altitude is chosen (subject to regulation), the lens and sensor are bought, the image width is the camera's native count. GSD is the mission's resolution contract — the number a client specifies ("2 cm GSD"), a regulator audits, and a surveyor defends — and it is computed, before a single battery is charged, by exactly the four fields above.
Worked Example: A 1-Inch Drone at 100 Metres
The defaults price the workhorse mapping setup: a 1-inch sensor (13.2 mm wide) behind an 8.8 mm lens, 5,472 pixels across, at 100 m. GSD = 13.2 × 100,000 ÷ (8.8 × 5,472) = 27.4 mm — 2.74 cm per pixel — and the footprint is 2.74 cm × 5,472 ≈ 150 metres across per frame (the similar-triangle shortcut agrees: 13.2 ÷ 8.8 × 100 = 150). At that GSD a car spans about 160 pixels and is unmistakable; a 10 cm crack in a runway spans under 4 pixels and is a rumour. The note's identification rule (4 pixels across for reliable detail) prices the mission's honesty: this configuration certifies features down to about 11 cm, and any promise finer than that must be bought with altitude, lens or sensor — the calculator's four levers, pulled before takeoff rather than apologised for after landing.
The Altitude Bargain
Halving altitude halves the GSD (finer detail) but also halves the footprint's width — quartering its area — so the same survey takes roughly four times the photographs, four times the flight lines, and a multiple of the batteries and processing hours. Doubling altitude buys coverage at the price of detail and, eventually, legality (most jurisdictions cap drone altitude near 120 m, which for the default camera sets a GSD floor around 3.3 cm at the ceiling — plan from the cap downward). This is why mapping specifications are negotiated in GSD: it is the single number where image quality, flight cost and air law meet, and why two quotes for "aerial photos of the site" can differ by an order of magnitude while both are honest — they are selling different centimetres per pixel. The calculator exists so the buyer can name the number, not the adjective.
The Oblique World Beyond the Nadir
Every figure on this page assumes the camera looks straight down (nadir), and the real world tilts: pitch the camera for an oblique inspection shot and the GSD stretches across the frame — pixels near the horizon cover metres while those at the bottom cover centimetres, the similar triangles shearing into trapezoids. Inspection pilots live in that gradient (a tower's top, farther from a orbiting drone, is sampled coarser than its base), and serious deliverables quote the worst GSD in frame. The calculator's clean number is the nadir contract — the baseline from which oblique work is discounted — and knowing the exact baseline is what makes the discount computable rather than impressionistic. Mapping's whole relationship with geometry is in that sentence: idealise precisely, then price the departure. The nadir figure is also the number that belongs in contracts and deliverable metadata, with the oblique discount stated as a separate, negotiated line — the difference between a specification and a surprise.
From Film Cameras to Focal Planes: A Short Lineage
GSD thinking predates drones by half a century: aerial survey from manned aircraft quoted photo scales (1:10,000 and the like), which are the same ratio in a cartographer's dialect — scale's denominator is altitude ÷ focal length, and GSD is that scale multiplied by the pixel's physical size. The digital era merely made the number per-pixel explicit and gave it an acronym. The continuity is reassuring for a buyer: the survey firm quoting 2.74 cm GSD and the 1960s photogrammetrist quoting a scale are selling the same triangle, and this page's formula — which a film-era planner would recognise instantly with the sensor width swapped for film width — is the translation. Technologies rebrand; similar triangles are forever, and the planner who learned the film era's scale arithmetic can audit a drone quote on sight, because the acronym changed and the geometry did not.
Related Tools
More from the Other Calculators hub: the Camera Field of View Calculator — the same triangle, read as angles — and the Focal Length Calculator for choosing the lens in this formula.
Frequently Asked Questions
What GSD do I need for my project, honestly?
By the identification rule — features need roughly 3–5 pixels — work backwards from your smallest feature: roof inspections and solar-panel audits want 1–2 cm (a cell crack is centimetres); construction progress and stockpile volumes are comfortable at 3–5 cm; agricultural stand counts and broad crop health run 5–10 cm; regional land-cover is happy at 25 cm and beyond. When a client cannot name a feature size, the honest sequence is feature first, GSD second, altitude third — this page computes the third from the second, and forces the first conversation to happen in the open.
Is GSD the same as image resolution or accuracy, then?
No, and the distinction is a profession's honour: GSD is sampling (the grid's spacing), resolution is what the lens and motion actually resolve (rarely finer than 1.5–2× the GSD once blur and optics vote), and accuracy is where a point's coordinates land against the Earth — governed by GPS, ground control and the photogrammetry, often several GSDs in size. A map can be finely sampled and badly located. The calculator computes the sampling honestly and labels it as such; vendors who quote GSD as "accuracy" are selling the grid as the geography, and this page's precision about the difference is deliberate.
Why must I use the real focal length, not the 35mm equivalent figure?
Because the triangle is physical: the sensor's real millimetres pair with the lens's real millimetres, and the "equivalent" focal length is a field-of-view translation for a different (full-frame) sensor that would corrupt the ratio. A drone advertised as having a "24mm equivalent" lens may have a true focal length of 8.8 mm behind its 1-inch sensor — the defaults here are exactly that common case. Manufacturer spec sheets list the true figure (sometimes as "actual focal length"); the calculator's label insists on it because the error is not small — using 24 for 8.8 would misstate the GSD nearly threefold.
How does overlap planning use the footprint figure?
Photogrammetry needs each point seen in many images, so flight lines are spaced at a fraction of the footprint: at the standard 75% forward overlap and 65% side overlap, consecutive photos advance by 25% of the footprint's length and lines step by 35% of its width — the calculator's 150 m width implies lines about 52 m apart for this camera. The footprint figure is thus the mission's second budget: divide the site by the effective step grid and the photo count (and flight time, and processing load) falls out. GSD gets the glory; the footprint does the logistics; the same four inputs produce both.
Does terrain height change the GSD?
Continuously: the formula's altitude is height above the ground in frame, so over hilly terrain the GSD breathes — finer on the ridges (closer to the camera), coarser in the valleys — and a mission specified at 100 m over a site with 30 m of relief delivers GSDs varying by ±15% or so across the map. Terrain-following flight modes exist to hold the altitude-over-ground (and hence the GSD) constant, and specifications for legal or engineering work are written against the worst (coarsest) GSD in the block. The calculator computes the flat-ground figure exactly; treat relief as a margin to plan against it, and quote clients the coarsest figure the terrain will produce rather than the finest the launch point enjoys — the difference being where surveying reputations are kept.
What about satellite imagery GSD?
The same concept at orbital scale, quoted in metres: the best commercial satellites sample near 0.3 m per pixel, workhorse constellations at 3–10 m, and the free Landsat/Sentinel workhorses at 10–30 m — the identification rule then reads nations instead of runways (at 10 m, buildings are pixels and roads are lines; at 0.3 m, cars are objects). The formula is altitude's revenge: from 500 km, even enormous optics sample coarsely, and this page's four levers — with altitude in kilometres and focal lengths in metres — describe a telescope's mission as faithfully as a drone's, which is the mark of a real formula. It is also why spy-satellite legends are told in GSD: the number is the capability, stripped of adjectives.