Integrating Drone Orthomosaics and Point-Cloud Data

Integrating Drone Orthomosaics and Point-Cloud Data

Drone-derived data has become a standard deliverable in modern GIS and surveying workflows. Two products appear most often: the orthomosaic, a geometrically corrected, mosaicked aerial image, and the point cloud, a dense set of three-dimensional coordinate measurements produced by photogrammetry or LiDAR. This lab teaches you to bring both products into ArcGIS Pro, verify their accuracy, classify and derive terrain surfaces, and assemble a finished cartographic layout.

Understanding the Source Data

An orthomosaic is delivered as a GeoTIFF. Because perspective distortion has been removed and the image has been tied to ground control, features can be measured directly on it. The accuracy of that tie depends on the quality of the ground control points (GCPs) used during the photogrammetric processing. GCPs are surveyed with a GNSS receiver capable of centimeter-level accuracy; their coordinates are written into the processing software so the model is anchored to a known datum and coordinate reference system (CRS).

Point clouds from drone photogrammetry are stored in LAS or LAZ format (LAZ is the compressed version). Each point carries X, Y, Z coordinates and may carry additional attributes such as intensity, return number, and classification codes. The American Society for Photogrammetry and Remote Sensing (ASPRS) defines standard LAS classification codes: code 2 = ground, code 3–5 = vegetation (low, medium, high), code 6 = building, and others. ArcGIS Pro reads these files natively, either as a local LAS dataset or as a Scene Layer Package (SLPK) for web-optimized display.

Step 1 – Georeferencing Verification

Before you use any drone product in a professional deliverable, you must verify its positional accuracy. Open the GNSS control-point CSV in ArcGIS Pro and display it as an XY event layer. Compare each GCP's surveyed coordinate with the coordinate reported by the photogrammetric software at the same point. The difference is the residual error. Root mean square error (RMSE) computed across all GCPs is the standard accuracy metric. For mapping-grade work, a common target is that the RMSE does not exceed one to two times the orthomosaic's ground sample distance (GSD) — the real-world size of one pixel. Document your residuals in a table; note whether they exceed the GSD threshold and state the consequence for map accuracy. A large residual means features are displaced; parcel boundaries derived from the image would not match surveyed deed lines.

Step 2 – Point-Cloud Classification and Bare-Earth DEM

In ArcGIS Pro, add your LAS or LAZ file as a LAS dataset. Open the 3D Scene Viewer. Use the Layer Properties > Symbology panel to display points by classification code. If the file arrived unclassified, you can run the Classify LAS Ground and Classify LAS by Feature geoprocessing tools to assign ground, vegetation, and structure classes automatically.

Once ground points (class 2) are isolated, run LAS Dataset to Raster (or Create Terrain then surface to raster) to generate a bare-earth digital elevation model (DEM). Choose an appropriate cell size — typically equal to or slightly larger than your average point spacing. Apply the Hillshade renderer to the DEM in the Map Properties. A hillshade simulates illumination from a defined azimuth and altitude, making terrain relief visible beneath the orthomosaic. In the 2D layout, place the hillshade as the lowest layer, set the orthomosaic above it at reduced opacity (commonly 60–80%), and confirm that terrain texture shows through without obscuring image detail.

Step 3 – Cadastral Overlay and Cartographic Finishing

Add the parcel shapefile. Style boundaries with a thin, high-contrast line (bright yellow or white works well over aerial imagery) and set the fill to no color so the orthomosaic beneath remains visible. Adjust the orthomosaic's transparency until parcel lines and image features are both legible — this is a judgment call you refine visually.

Every professional map must include a data-source citation in the margin. For drone-derived layers, the citation must state: the collection date, the GSD of the orthomosaic, the coordinate reference system (horizontal datum and projection), and the name of the data provider or operator. Parcel data should cite its source agency and publication date. These fields are not optional; they allow a map reader to assess currency and accuracy independently.

Coordinate Systems and Datum Awareness

All layers must share the same CRS before they can be compared or published together. Confirm that the orthomosaic, point cloud, and parcel shapefile are registered to the same horizontal datum (commonly NAD83 in the United States) and the same projected coordinate system. ArcGIS Pro will project layers on the fly, but relying on that for final products can hide small datum shifts. Reproject any layer that does not match the project CRS natively.

Safety and Data Integrity Considerations

Drone data collection is covered by FAA Part 107 rules; you are working with the post-processed products in this lab, but you should always confirm the flight was conducted legally and that the data you receive has a documented chain of custody. When handling field GNSS equipment for GCP verification outside this lab, wear appropriate high-visibility PPE in traffic or near active machinery, follow your organization's field safety plan, and never approach a drone's propeller arc while power is applied. In the software environment, save your ArcGIS Pro project frequently and keep an unmodified copy of the raw LAS file — point-cloud processing is destructive to classification attributes if original files are overwritten.

Assembling the Final Layout

In the ArcGIS Pro Layout view, add the standard map elements: title, north arrow, scale bar, legend, and the data-source citation block. Confirm the legend correctly identifies all three data layers. Export as a PDF or high-resolution TIFF for submission. The instructor will evaluate whether the hillshade–orthomosaic composite renders terrain clearly, parcel boundaries are distinguishable, and all required metadata fields appear in the citation block.