I run about 30 roof survey flights a year in the South Bay, and the most common question from roofers receiving my deliverables is: what am I supposed to do with this orthomosaic file? An orthomosaic is a processed image product, not a raw photo, and the workflow that produces it is not obvious from looking at the output. Once a roofer understands how to read one, it becomes the most useful single document in a roof condition assessment: geometrically correct, measurable, and showing the whole roof in one image at full resolution.
What an orthomosaic is and why it matters for roof work
A standard aerial photo of a roof, taken from a single camera position, is a perspective image: elements near the camera appear larger than elements far away, and the roof surface is distorted by the camera’s angle relative to it. An orthomosaic corrects for these distortions by stitching together hundreds of overlapping nadir images — straight-down shots from a consistent altitude — and applying geometric corrections to produce a single image in which every point on the roof is at its correct scale relative to every other point.
You can measure from an orthomosaic: ridge-to-eave run, valley length, hip length, penetration spacing — all taken off with a calibrated scale bar or the built-in measurement tools in most photogrammetry software. For a roofer quoting a reroof, that replaces a manual pitch measurement from inside an attic. For an adjuster calculating damaged area, it replaces a ground estimate. The FAA’s Part 107 guidance permits commercial aerial data collection including photogrammetric surveys; any drone operator generating deliverables for roofers or insurers should operate under a current Part 107 certificate.
The flight parameters that produce a usable orthomosaic
Orthomosaic quality depends primarily on three flight parameters: altitude, overlap, and camera angle. I fly roof surveys at 80 to 120 feet above the surface for the nadir grid that produces the orthomosaic. Lower altitudes produce higher resolution but require more passes. Higher altitudes cover more area per image but lose detail in the granule texture of composition shingles.
Overlap is critical. My standard is 80 percent frontal and 70 percent lateral, meaning each roof point appears in approximately 20 to 25 frames. That redundancy lets the photogrammetry software reconstruct accurate geometry. At 70 and 60 percent — the minimum most software accepts — gaps or alignment errors appear at complex geometries: valleys, hips, and areas under tree canopy.
Camera angle for the orthomosaic grid is nadir only. I do not mix oblique frames into the nadir grid because the software handles them differently and mixing produces geometry errors. Oblique passes for shingle condition, valley debris, and penetration flashing are a separate flight at a different altitude, in a separate folder from the grid frames.
Reading the orthomosaic for roof condition
The orthomosaic shows two kinds of information: geometry and surface condition. Geometry is the easy read — slope layout, ridge and valley positions, penetration locations, all accurate and measurable. Surface condition is more nuanced: granule loss appears as lighter patches, lifted shingles as shadows along their uphill edge, moss growth as dark texture with different reflectance, and replaced sections as a different color zone.
What the orthomosaic does not show: sub-shingle damage, flashing condition at vertical penetrations, and damage on the underside of lifted shingles. For these, oblique close-up passes are necessary. An orthomosaic without oblique coverage tells you the overall pattern; oblique coverage tells you the specific failure modes. Both are in every survey I deliver.
Delivering the orthomosaic to a roofer
My standard delivery is: the full-resolution orthomosaic as a JPEG or PNG with a scale bar in the corner, an annotated version with damage zones outlined and labeled, and the GeoTIFF for any client with GIS or roofing-specific measurement software. I also include a panel-area spreadsheet — each roof plane by name with real-world area in square feet — which saves the roofer 30 to 45 minutes of manual measurement on a complex hip roof. The notes on drone progress shots on a La Mesa hillside hardscape cover repeated flight patterns, GPS waypoint repeatability, and reading consecutive aerial sets.
Frequently asked questions
What is the resolution of a drone orthomosaic for a residential roof? Resolution depends on altitude and camera. At 80 feet with a 1-inch sensor camera, a typical residential roof orthomosaic has a ground resolution of 0.4 to 0.8 inches per pixel. That is sufficient to identify individual shingle tabs, granule loss zones, and flashing position, but not to read manufacturer labels on materials or assess shingle tab surface texture at the microscopic level.
Why do I need 80 percent overlap for a roof orthomosaic? High overlap means each point on the roof appears in many frames from slightly different angles, which gives the photogrammetry software enough information to reconstruct accurate geometry. At lower overlap, complex roof shapes — steep pitches, valleys, areas in shadow — produce reconstruction errors that appear as distortions or gaps in the finished orthomosaic. For a flat or low-pitch roof, 70 percent overlap is sometimes adequate; for a complex pitched roof, 80 percent is safer.
Can a roofer use an orthomosaic to generate a material estimate without visiting the site? Yes, for the area calculation portion. The orthomosaic gives accurate roof plane areas for the material takeoff. However, the on-site visit is still necessary to assess the condition of the decking, the ventilation, and the penetration flashings that the drone cannot reach. Think of the orthomosaic as replacing the tape measure and the pitch gauge, not the site visit itself.