RTK, PPK and GCPs in Drone Surveys
Understand RTK, PPK, ground-control points, checkpoints and accuracy specifications before commissioning a drone mapping project.
By SkyViews Aerial Drone Studio
Survey clients often hear RTK, PPK, GCP and centimetre accuracy used together as if they mean the same thing. They do not. The right workflow depends on the site, equipment, control, conditions, processing and the use of the final data.
RTK and PPK in simple terms
RTK can use corrections during capture. PPK applies positioning corrections during processing. Both can improve georeferencing, but neither automatically guarantees a particular final accuracy in every environment.
What ground-control points do
GCPs are surveyed points visible in the imagery that help constrain a mapping project. Checkpoints are independent points used to test the result. Their number, distribution, survey quality and visibility matter more than simply saying “GCP used.”
Specify accuracy by use
Ask whether the output is for visual planning, stockpile volume, earthwork, design coordination, GIS, engineering review or legal boundary work. Define horizontal and vertical tolerance, coordinate reference, units, control method, checkpoint reporting, surface and exclusions.
Check the deliverable
Request the orthomosaic, point cloud, surface model, contours or mesh in the agreed format together with coordinate information, quality statement, control/checkpoint results where scoped, capture date and limitations. Read the mapping QA checklist before acceptance.
Start with the decision, not an accuracy slogan
A client may hear “RTK drone” and expect every point in a map to be correct to a few centimetres. RTK describes one positioning method during capture. It does not tell the client how well the finished orthomosaic, ground model or volume calculation matches the site. Begin with the intended use. A visual planning map, a stockpile estimate and a design surface need different checks. Write the acceptable horizontal and vertical error for the particular output, or ask a survey professional to define it.
State whether the result will guide a meeting, support engineering design, feed GIS, compare earthwork or inform a legal boundary decision. For the last use, imagery alone is not enough; records and the responsible survey process matter. The proposal should name the final product and its acceptance test. A supplier cannot prove the whole dataset meets a requirement simply by showing a controller screen with a “fixed” position during flight.
Understand RTK, PPK and control in one workflow
RTK applies positioning corrections while the aircraft is flying. PPK applies corrections afterward using recorded observations. Both can improve camera positions, but the result still depends on a suitable base or correction source, reliable observations, camera calibration, image overlap and processing. If the correction was lost for some photos, the team should identify those images and explain how they were handled. A processing report is more useful than a badge saying “RTK enabled.”
Ground-control points are marked features whose positions are measured separately and used to tie the image block to the project coordinates. Checkpoints are measured independently and held out of that fitting step to test the final result. Ask which points served which role. A point cannot fairly prove a model's independent accuracy if it was also used to force the model into place. The USGS imagery-calibration guide explains why control quality and image geometry matter even when the pixels are very small.
Ask where the reference coordinates came from
Every result needs a coordinate reference and units. Ask whether the project uses a local grid, a national or global reference, and what height reference is intended. A model may look aligned on screen while using a different datum from the design drawing. That mismatch can be more damaging than a small measurement error. Record the base station, correction source, survey method and any transformation into the client's project grid.
If a contractor provides existing control points, the surveyor should check their condition and suitability. A painted mark may have moved or been covered. A point measured in one reference system should not be imported into another by changing the label. The receiving CAD or GIS team should confirm sample coordinates and heights before using the full dataset. Put the coordinate information in the delivery note and metadata, not only in an email.
Place checks where errors could hide
Control and checkpoints should cover the project's extent and changes in elevation where practical. If all points sit near one gate, they say little about the far edge of a long site. Place checks on stable, visible ground, away from moving stockpiles or shadows that make a target hard to mark in imagery. Do not ask a crew to enter an unsafe slope just to spread points evenly. Record areas where a good check could not be established and explain the effect on confidence.
Check the final product, not only the camera positions. For an orthomosaic, compare visible checkpoint locations. For a terrain model, compare independent ground elevations at suitable surfaces. A dense canopy, water or glass roof may prevent the camera from seeing the ground at all. RTK or PPK cannot recover a hidden surface. A ground crew or other sensor may need to fill those gaps.
Read an accuracy report in plain language
A useful report lists the specification, coordinate system, check-point method, number and spread of independent points, horizontal and vertical differences, summary statistics and problem areas. A single average can hide a poor corner of the site. Ask for the largest errors and a map of check locations. If the result failed the agreed target, ask whether it can be improved, whether the affected area must be resurveyed or whether the intended use should be narrowed.
Ground sample distance, often called GSD, is the size of a pixel on the ground. It helps describe visible detail; it is not the same as positional accuracy. The USGS explicitly warns that a small GSD does not make the dataset equally accurate (USGS guidance). A crisp roof edge can still be in the wrong coordinate position. A pretty orthomosaic is not a substitute for tested checkpoints.
Keep horizontal, vertical and relative quality distinct
Horizontal accuracy concerns map position. Vertical accuracy concerns height. Relative quality describes how consistently nearby features fit together. A project may be acceptable for a marketing site plan but poor for cut-and-fill calculations if its heights are weak. Ask which dimension matters for the decision. For a stockpile volume, the top surface, base surface and boundary all matter. For a road alignment, position and height may both be critical. For a simple visual progress record, a formal height surface may not be needed.
Do not treat a percentage printed by software as an accuracy statement without knowing what it measures. A processing “quality” score may describe image matching, not independent map accuracy. The survey lead should translate software output into the project's acceptance criteria and document any limitations. If the client has a standard, name it in the scope and say exactly which part is being applied.
Example: a construction earthwork survey
Suppose a developer needs a monthly surface model of an open construction site. The surveyor defines the project grid and a vertical reference, marks control points and reserves separate checkpoints. The pilot captures overlapping images with RTK corrections and records the base source. The processing team builds a surface model and checks it against the independent ground points. One vegetated corner has weak ground visibility, so the report excludes it from the earthwork calculation and asks for ground infill.
The next month, the team repeats the method. Both surfaces use the same reference and include their tested check results. The engineer can now review a change volume within the covered area, while the report keeps the vegetated gap visible. A claim of “centimetre-accurate drone survey” would have hidden the most important caveat; the evidence-based report explains exactly what can be used.
Questions for a mapping provider
Ask what output will be delivered, what coordinate and height references it uses, how RTK or PPK corrections are obtained, whether ground control is needed, how independent checks are spread and how failed areas will be reported. Request a sample quality statement before booking. Agree on the intended use and who signs off the result. The mapping QA checklist can be used at handover, and the survey comparison guide explains when ground measurements are still the simpler choice.
FAQs
Does RTK mean the map is survey-grade?
No. RTK is one part of the workflow. Site conditions, control, processing, checkpoints and the stated specification determine whether the result suits the intended use.
Are GCPs always necessary?
Not for every project. The decision depends on the equipment, site, use, accuracy target and independent quality-check requirement.
Can a drone replace a legal boundary survey?
Not automatically. Boundary and legal-property decisions may require a licensed surveyor and a separately specified method.
Plan your project
Share your site, dates and intended output so we can confirm a suitable scope.
Drone Survey & Mapping · Compare project scope · Get in touch
Related reading
- Drone Roof and Facade Surveys for Facility Maintenance Teams
- Drone Mapping File Formats Explained: Orthomosaics, DSM, DTM, LAS and CAD
- Drone Photo Metadata and Geotagging: What Indian Clients Should Receive
- Drone Video Post-Production in Bangalore: Editing, Colour and Delivery Guide
- Indoor FPV Drone Filming in Bangalore: Production and Safety Guide
- Restaurant, Café and Venue Drone Photography in Bangalore
Contact SkyViews Aerial Drone Studio at +91 90190 56479 or skyviewsdronestudio@gmail.com. Service area: Bangalore, Karnataka and India and wider Bangalore, Karnataka and India.