Quick Answer
Ground control points (GCPs) are physical markers placed on the ground with precisely known coordinates. They give drone mapping software fixed reference points to anchor aerial images to real-world positions, improving map accuracy from metres down to centimetres. Whether you are flying a standard drone or one with RTK correction, GCPs remain the most reliable way to verify and improve survey-grade results.
What Are Ground Control Points?
A ground control point is a marked location on the earth whose geographic coordinates have been measured with high-precision survey equipment, typically an RTK or PPK GNSS receiver. When a drone flies over the site, its camera captures these markers in the aerial imagery. During processing, the mapping software uses the known coordinates of each GCP to georeference the entire dataset, correcting positional errors and distortion across the map.
Without GCPs, a drone map relies solely on the GPS module onboard the aircraft. Consumer-grade GPS is accurate to roughly 1 to 2 metres, which is acceptable for casual photography but nowhere near precise enough for surveying, construction, or land management work. GCPs reduce that positional error to centimetre-level by anchoring the processed map to independently verified points on the ground.
Why GCPs Matter for Accuracy
Correcting Systematic Errors
Drone imagery suffers from several distortion sources: imperfect camera lens calibration, barometric altitude errors, and the inherent imprecision of standard GPS. Even small individual errors compound across hundreds of overlapping photos. GCPs give the photogrammetry engine hard reference data to correct these errors systematically across the entire map.
Absolute vs Relative Accuracy
A drone map can look visually correct in terms of relative accuracy, meaning distances between features in the map match real-world measurements, while still being offset from its true position on the globe. GCPs establish absolute accuracy, ensuring the map sits at the correct real-world coordinates. This distinction matters when you need to overlay drone data with existing GIS datasets, cadastral records, or engineering plans.
Quality Assurance
Beyond improving accuracy, GCPs provide a quantifiable measure of map quality. By comparing processed map coordinates against known ground coordinates at checkpoint locations (separate points not used in the adjustment), surveyors can calculate root-mean-square error (RMSE) values and produce an accuracy report. This is often required for contractual compliance on commercial survey projects.
Do You Still Need GCPs with an RTK Drone?
This is one of the most common questions in drone surveying. RTK (Real-Time Kinematic) drones receive live correction data that positions each photo to centimetre-level accuracy during flight. PPK (Post-Processing Kinematic) achieves the same result by correcting positions after the flight using logged satellite data.
The short answer: RTK dramatically reduces your reliance on GCPs, but it does not eliminate them entirely. Here is why.
RTK correction depends on maintaining a stable radio link between the base station and the drone. Signal interruptions caused by terrain, structures, or electromagnetic interference can produce positional errors that go undetected until processing. A small number of GCPs (three to five) spread across the site acts as an independent check on the RTK solution, catching any drift or ambiguity that slipped through.
For many projects, a hybrid approach works well: use RTK for the drone and place a reduced number of GCPs for verification. For the highest-stakes work, such as legal boundary surveys or construction set-out, full GCP coverage alongside RTK gives the most defensible results.
Types of Ground Control Points
Passive Targets
Passive GCPs are the most widely used type. They are physical markers placed on the ground before the flight, designed to be clearly visible in aerial photographs. Common forms include:
- Checkerboard pads and< these are square tiles with alternating black-and-white squares and a clearly defined centre point. They are the gold standard because the centre is unambiguous, even at high flight altitudes.
- Spray-painted marks, typically an "L" shape where the inside corner marks the precise coordinate. An "L" is preferred over an "X" because the intersection of two painted lines can span several centimetres, introducing selection error during processing.
- Pre-made GCP targets, manufactured from durable plastic or fabric with a matte finish to prevent glare. These are portable, reusable, and available in high-contrast colour combinations such as black-and-white or orange-and-black.
Natural Features
In some situations, surveyors use existing features as control points: a corner of a paved road, the edge of a manhole cover, or a distinct rock formation. These require no setup time, but they come with significant drawbacks. Shadows, low contrast, and ambiguity about the exact measurement point make natural features less reliable than purpose-made targets. Use them only when access or time constraints prevent placing proper markers.
Active Targets
Active GCPs emit a signal (often retroreflective or electronic) that can be detected automatically by processing software. These are less common in general drone mapping but are used in specialised survey workflows where automated target detection speeds up processing.
GCPs vs Checkpoints
The distinction between GCPs and checkpoints is one of the most misunderstood aspects of drone surveying, yet it is straightforward.
GCPs are entered into the photogrammetry software as known reference points. The software uses them to adjust and warp the map so that it aligns with real-world coordinates. GCPs directly influence the final map output.
Checkpoints are measured with the same survey-grade equipment and placed in the same way, but they are not included in the processing adjustment. Instead, after the map is generated, the checkpoint locations in the processed map are compared against their known ground coordinates. The difference tells you how accurate the map actually is.
Physically, GCPs and checkpoints look identical. The difference is purely in how the data is used during processing. A best practice is to designate roughly 20 to 30 percent of your ground points as checkpoints and the remainder as GCPs.
How Many GCPs Do You Need?
The number of ground control points depends on three factors: site size, terrain complexity, and the accuracy standard you need to meet.
Research by the Nevada Department of Transport, published in partnership with Pix4D, found that accuracy improves significantly up to about 5 GCPs, with diminishing returns beyond 10. For most projects, 5 to 10 GCPs provides an optimal balance between field time and map quality.
Guidelines by Project Type
- Small site, flat terrain, standard drone: 5 to 8 GCPs for areas up to 20 acres.
- Medium site, moderate terrain, RTK drone: 3 to 5 GCPs for areas up to 50 acres.
- Large site, complex terrain: 8 to 15 GCPs, with spacing of roughly 200 to 400 metres between points.
- High-precision cadastral work (1.5 cm resolution): GCPs every 100 to 200 metres.
- General topographic mapping (3 to 5 cm resolution): GCPs every 300 to 500 metres.
For undulating or complex terrain, increase the number by 10 to 20 percent to account for the additional elevation variation that can introduce distortion.
How to Place GCPs
Correct placement is as important as the number of points. Poorly distributed GCPs can actually degrade map quality rather than improve it.
Distribution
GCPs should be spread across the entire project area, not clustered in one section. A common mistake is concentrating points around the most important feature, which leaves the rest of the site poorly controlled and can cause warping at the edges.
Ideally, place GCPs near the corners of the site boundary with additional points along the edges and at least one near the centre. If the site is irregularly shaped, ensure every major section has coverage.
Elevation Considerations
Always place at least one GCP at the highest accessible point on the site and one at the lowest. This captures the full range of elevation and helps the photogrammetry engine perform accurate vertical projection. Skipping this step can produce systematic height errors across the entire model.
Edge Coverage
The outermost GCPs should define a boundary that encompasses your area of interest. Avoid placing all GCPs inside the mapped area with none near the edges, because accuracy degrades rapidly outside the convex hull defined by the control points. Extending GCP coverage slightly beyond the area of interest ensures edge accuracy.
Visibility and Stability
Each GCP must be clearly visible in the aerial imagery. Avoid locations near tall structures that cast shadows, areas of dense vegetation, or spots where the marker could be obscured by parked vehicles or equipment. The marker should sit flat on stable ground, not on loose gravel or soft soil where it might shift between placement and the flight.
How to Survey GCP Coordinates
Once the physical markers are in place, each one must be surveyed with centimetre-grade GNSS equipment. This is where an RTK GNSS receiver like the Emlid Reach RS4 Pro or the Emlid Reach RS4 Pro and RS4 Base/Rover Kit comes in. These receivers deliver survey-grade positioning as either a base station transmitting corrections or a rover collecting point data.
The workflow is straightforward:
- Set up your base station over a known point, or establish a new one by averaging a static position.
- Connect the rover to receive corrections via UHF radio or NTRIP over cellular.
- Walk to each GCP marker, place the rover pole tip on the centre point, and log the coordinate using a survey app such as Emlid Flow.
- Record each point with a unique identifier that matches the label you will use in your photogrammetry software.
For post-processing workflows, you can also log raw observation data at each point and apply PPK corrections later using software like Emlid Studio. This removes the need for a real-time radio link and can be more reliable in areas with poor cellular coverage.
Common Mistakes to Avoid
Using an "X" Pattern for Spray-Painted GCPs
An "X" made with spray paint creates an intersection zone that can be several centimetres wide. At high resolutions (sub-2 cm per pixel), this introduces a projection error that feeds directly into your final map accuracy. Use an "L" shape instead, where the inside corner marks a single, unambiguous point.
Clustering GCPs in One Area
Concentrating all your control points near the area you care about most, while ignoring the rest of the site, is counterproductive. The photogrammetry engine has insufficient reference data for the outer areas, leading to distortion that can propagate inward and affect the clustered zone as well.
Forgetting to Record Point Identifiers
Each GCP needs a unique label that is both written on or near the marker and recorded in your survey data. If you cannot match a point in the aerial imagery to its coordinate record, that GCP is useless. Number or letter each target clearly and keep a field log.
Placing GCPs on Unstable Ground
A GCP on soft mud, loose gravel, or a surface that shifts between placement and the flight will have moved by the time the drone captures it. The coordinate you logged no longer matches the marker position in the imagery, introducing error rather than correcting it.
GCP Best Practices Summary
- Use 5 to 10 GCPs for most projects, with additional checkpoints for verification.
- Distribute points across the entire site, including corners, edges, and both high and low elevations.
- Use high-contrast, matte-finish markers with a clearly defined centre point.
- Survey each GCP with RTK or PPK GNSS equipment for centimetre-grade coordinates.
- Even with an RTK drone, include at least 3 to 5 GCPs as a quality control check.
- Keep field notes mapping each GCP label to its location for reliable processing.
- Extend GCP coverage slightly beyond the area of interest to maintain edge accuracy.
Getting Started
If you are setting up a drone mapping workflow and need reliable GNSS equipment for GCP collection, the Emlid Reach RS4 Base Station Kit gives you everything needed to establish a base and collect survey-grade control points. For a complete base-and-rover setup, the RS4 Pro and RS4 Base/Rover Kit covers both roles in one package.
Browse our full range of GPS and navigation systems for drone mapping and surveying equipment. For more on the underlying positioning technology, see our guide to RTK vs PPK for Drone Mapping. And for help getting your receiver configured, our Emlid Reach Base and Rover Setup Guide walks through the process step by step.
FAQ
Can I use a smartphone GPS to survey GCP coordinates?
No. Smartphone GPS is typically accurate to 3 to 5 metres under good conditions, which is worse than the error you are trying to correct. GCPs must be surveyed with RTK or PPK GNSS equipment capable of centimetre-level accuracy.
What colour should GCP markers be?
High-contrast combinations work best: black and white, orange and black, or bright pink and green. The marker must stand out clearly from the surrounding terrain in the aerial imagery. Matte finishes are preferable to glossy ones, which can reflect sunlight and obscure the centre point.
How long does it take to place and survey GCPs?
For a typical 10-acre site with 5 to 8 GCPs, allow 30 to 60 minutes for placement and another 30 to 45 minutes for surveying coordinates with an RTK rover. Total field time is usually 1 to 2 hours.
Do I need GCPs if my drone has PPK?
PPK significantly reduces, but does not eliminate, the need for GCPs. The same principle applies as with RTK: use a reduced number of GCPs (3 to 5) as independent verification, especially on larger or topographically complex sites.
What happens if I do not use any GCPs?
With a standard (non-RTK) drone, your map will have positional errors of 1 to 5 metres or more. With an RTK drone, accuracy improves to a few centimetres, but you have no independent way to verify that accuracy. For any project where measurement precision matters, GCPs (or at minimum checkpoints) are non-negotiable.