Drone Mapping QA Checklist: Before You Accept an Orthomosaic or 3D
A client-side quality checklist for reviewing drone mapping deliverables, control, coverage, coordinate systems, exports and documented...
By SkyViews Aerial Drone Studio
A mapping deliverable should be reviewed against the original brief, not judged only by how sharp the aerial image looks. This checklist helps landowners, developers, contractors and survey teams ask the right questions before accepting an orthomosaic, point cloud, surface model, contour set or 3D mesh.
Confirm the brief
Check the site boundary, capture date, intended use, requested coordinate reference system, ground sampling or accuracy target, control requirements, output formats and whether the work is for planning, measurement, design coordination, marketing or record keeping.
Write the decision at the top of the review sheet. “We need a current image of the whole site for planning” is a different request from “we will calculate stockpile volume for payment.” The second needs a stated measurement method, reference surface and quality check. If the brief says only “3D map,” ask the team which file and accuracy the receiving software needs. An attractive screenshot is not enough to settle that question.
Check the exact boundary used for capture. Was it taken from a client drawing, a site walk, a legal plan or an approximate pin on a map? The image should not be presented as a legal boundary simply because a red line is overlaid on it. Label the source of every boundary line and ask who is responsible for its correctness. If different teams supplied different boundaries, resolve that before judging whether the flight covered the right area.
Check coverage
Confirm that the delivered area matches the agreed boundary and that important edges, access roads, stockpiles, structures and ground features are visible. Look for gaps caused by vegetation, shadow, reflective surfaces, moving equipment or restricted flight paths.
Inspect the whole output at a useful zoom level, then inspect key work areas closely. It is easy to miss a small blank strip along a fence or under a tree when viewing a reduced preview. Check the edges of the site, narrow access ways, corners of a roof and any location where a quantity or design decision will be made. Mark gaps on a copy of the map and ask whether they were expected, excluded or missed.
A dark shadow may still show enough for a visual record but be unsuitable for picking a clear edge. Water and glass can cause matching errors. A crane, vehicle or person that moved between images may appear twice or create a warped patch. The reviewer should ask for the original images when an important feature looks strange. In some cases a different flight time or a small ground survey will solve the gap better than more processing.
Review control and reference information
Ask whether ground control points, checkpoints, RTK/PPK data or another reference method was used. Confirm the coordinate system, vertical datum and units. A file with coordinates but no documented reference system can be difficult to use safely in CAD or GIS.
Ask the provider to explain which points helped build the model and which independent points were used to check it. A control point and a checkpoint do different jobs; using the same point for both does not give an independent check. The report should identify the points, their source, how they were measured and the results that matter for the agreed use. Do not accept a single “centimetre accuracy” line without method and evidence.
Open the data in the receiving team's software and compare a few known positions. A map can look right in one viewer yet appear shifted when another team uses a different coordinate reference or vertical datum. Confirm whether heights use metres, what zero level they refer to and whether a local site grid was applied. If the deliverable will be joined with CAD drawings, ask the CAD user to test one known feature before signing off. Catching a shift during handover is easier than finding it after design work has started.
Inspect the outputs
Review the orthomosaic for seams, warped buildings, duplicated objects, blurry areas and colour inconsistency. Review the point cloud or model for holes, noise, vegetation artefacts and missing vertical faces. Open the actual delivered formats in the software your team will use.
For an orthomosaic, check the road edges, roof corners and straight walls. These make obvious test features because a bad image join can bend or repeat them. A mosaic is made from many photographs; it is not a single photograph taken straight down. Tall objects can lean or stretch, especially near the edge of the capture. Decide whether that effect matters for the stated use. It may be acceptable for a broad planning view and unacceptable for a detailed asset outline.
For a point cloud, look at density and gaps as well as colour. Can you see the ground where the client expects to measure it, or only the tops of vegetation and buildings? A digital surface model represents visible surfaces. A terrain model needs an additional ground-selection method and review; it should not silently be treated as bare earth. If the deliverable includes a 3D mesh, turn it around and inspect walls, undersides and hidden corners. A model can look convincing from the hero angle while having no usable information on the opposite side.
Check export sizes and versions too. Large files may need tiling or a different transfer method. A PDF report should use the same site name, coordinate information and version as the data folder. If the map and report disagree, stop the acceptance review until the provider identifies which is current.
Read the quality statement
The report should state the method, capture conditions, processing software or workflow where relevant, control/checkpoint results if available, known limitations and any areas excluded from measurement. Do not infer centimetre-level accuracy from a marketing phrase alone.
Read limitations before using a measured output. Wind can move vegetation. Shiny surfaces can fail to match. A steep stockpile face can be hidden from overhead cameras. A road under tree cover may not appear at all. Ask how these areas were handled and whether a ground method was used to fill the gap. A useful quality statement says which parts are reliable for the agreed task, which are approximate and which are outside scope.
For a volume figure, ask what base surface was used and when it was measured. Two teams can compute different volumes from the same pile if they use different base lines. For a progress comparison, ask whether the earlier and later maps share a compatible coordinate system and check method. A difference image should not turn registration error into an apparent earthwork change.
Handover checklist
- Agreed boundary and capture date
- Orthomosaic and/or model in the requested format
- Point cloud, surface model or contours if included
- Coordinate reference system and units
- Control/checkpoint or accuracy documentation where scoped
- File naming, version and readme
- Clear list of exclusions and recommended follow-up
Add a contact name for technical questions and a simple readme that explains what each file contains. If the team may revisit the site, keep the flight date, ground-control records, naming system and any repeat-view plan together. A future surveyor can then decide whether the new capture is truly comparable with the old one.
Plan the next step
If the project needs design-grade or legal-boundary information, involve the appropriate licensed survey or engineering professional. SkyViews can scope drone survey and mapping, orthomosaic and 3D model deliverables and a practical data handoff for downstream teams.
Record pass, rework and exception decisions
For each agreed acceptance criterion, record the evidence and one result: pass, rework or accepted exception. A missing corner of the agreed coverage area may require rework; an inaccessible area already excluded in the scope may be a documented exception. Neither should be silently removed from the report.
An acceptance table can be simple. List the requirement, the file or report page that proves it, the reviewer, the result and the action. For example, “north boundary visible” may pass on the orthomosaic. “Point-cloud ground in tree-covered strip” may be an agreed exception with a note to use a separate ground survey. “CAD file opens at the agreed coordinates” needs an actual test by the CAD team, not an assurance from a thumbnail. This format lets project managers and technical reviewers discuss the same evidence.
Keep format-opening checks separate from positional accuracy checks. Ask the receiving team to sign off on its own intended use. Build the acceptance table into survey and mapping so quality review is planned before capture.
FAQs
Is a high-resolution orthomosaic automatically survey accurate?
No. Resolution and accuracy are different. Accuracy depends on the method, control, conditions, processing and stated quality checks.
What coordinate information should be delivered?
The agreed coordinate reference system, units and vertical reference should be documented clearly enough for the receiving team to use the files correctly.
Can drone mapping replace a boundary survey?
Not automatically. Legal-boundary and design 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
- Orthomosaic vs 3D Model: Which Drone Survey Output Do You Need?
- Drone Client Approval Checklist: Review Video, Photos and Mapping Before Release
- Agriculture Drone Mapping for Farms and Estates in Karnataka: Practical Guide
- Drone Photography for Architects: Portfolio, Project and Design Documentation Guide
- Drone Footage Licensing and Usage Rights in India: Client Planning Guide
- Drone Shoot Weather and Rescheduling Guide for Bangalore Projects
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.