Rock Robotic Photogrammetry Orthophoto Generation
Aug 05, 2025
By Alex Knoll, CTO
Purpose and Scope
Drone mapping projects increasingly rely on orthophotos—geometrically correct aerial images that provide a true top‑down view of a site. Orthophoto generation usually relies on photogrammetry, an imaging process in which overlapping photographs are aligned and orthorectified into a single mosaic. Traditional photogrammetry packages like Pix4D and DroneDeploy compute orthophotos by matching features across images and using GNSS/RTK data to approximate camera positions. Rock Robotic’s R3 Pro V2 LiDAR system introduces a novel workflow that fuses a co‑aligned camera with LiDAR data to generate high‑accuracy orthophotos directly in ROCK Desktop. This white paper outlines the photogrammetry process used in ROCK Desktop, explains how it differs from traditional software, and summarizes the benefits of this approach.
Background: Photogrammetry and LiDAR
Photogrammetry creates maps and 3‑D models from overlapping photos. It is cost‑effective and captures real‑world textures and colours but its accuracy depends on precisely aligning images and maintaining consistent overlaps. Photogrammetry can struggle when vegetation obscures the ground, during low‑light or night operations, or where camera positions are not accurately known. LiDAR (Light Detection and Ranging) uses laser pulses and the time‑of‑flight of the returning signal to measure distances. LiDAR collects data day or night, penetrates vegetation and provides dense point clouds with centimetre‑level elevation accuracy. In essence:
- Photogrammetry excels at colour and texture but is limited by camera positions and feature matching.
- LiDAR excels at accurate 3‑D geometry and can penetrate vegetation and low‑light conditionslearn.rockrobotic.com.
Traditional drone mapping relies on either photogrammetry or LiDAR separately. ROCK Robotic’s R3 Pro V2 and ROCK Ultra brings these technologies together by co‑aligning an RGB camera with a high‑accuracy LiDAR sensor and using both datasets to generate orthophotos.
Overview of ROCK Photogrammetry
ROCK Photogrammetry feature
With the R3, R3 Pro V2, and ROCK Ultra, Rock Robotic introduced ROCK Photogrammetry. This is a new photogrammetry technique that allows precise capturing of building corners and object locations quickly and accurately. ROCK Photogrammetry is available on the R3, R3 Pro V2 and ROCK Ultra LiDAR systems, enabling users of these platforms to generate orthophotos directly in ROCK Desktop.
Processing workflow in ROCK Desktop
Photogrammetric processing in ROCK Desktop integrates LiDAR and camera data. The workflow is as follows:
- Import raw LiDAR, GNSS and camera data. Users select the flight lines and adjust range gates to remove unwanted points.
- Select outputs. During output selection, there is an option to “Generate Orthomosaic”. This option is available only for the R3, R3 Pro V2, and ROCK Ultra; other units cannot generate orthophotos.
- Trajectory and colorization processing. ROCK Desktop processes GNSS and IMU data (Post‑Processing Kinematics) to create a precise trajectory and uses LiDAR points to colorize the point cloud. During this process, ROCK writes EXIF data (X‑, Y‑, Z‑coordinates) into each photo’s metadata, making the photos georeferenced. Generating EXIF metadata is crucial for photogrammetry because many software packages require georeferenced images.
- Orthomosaic generation. ROCK Desktop uses the 3‑D LiDAR point cloud and the precise camera trajectory to rectify the images. Because the camera is co‑aligned with the LiDAR, the algorithm can project each pixel onto the LiDAR surface, correcting distortions and generating a seamless orthophoto. The resulting orthophoto can be exported as a .tif file. Rock notes that orthophotos from separate projects cannot be merged and must be generated individually.
In essence, ROCK Photogrammetry doesn’t rely solely on feature‑matching between photos; instead it uses the LiDAR‑derived 3‑D geometry and precise GNSS/IMU trajectory to orthorectify the imagery. This is fundamentally different from traditional photogrammetry tools.
Limitations and considerations
Because ROCK Desktop projects each photo onto the LiDAR surface rather than warping images to fill gaps, it cannot manufacture data where no photograph was captured. Areas that were not covered by any photo will appear as blank pixels in the orthophoto. Surveyors should therefore plan flights with sufficient overlap and coverage to avoid gaps. This LiDAR‑anchored approach also means orthophotos from separate projects cannot be merged into a single mosaic.
How ROCK Desktop Photogrammetry Differs from Pix4D and DroneDeploy
Source data and rectification approach
| Feature/Capability | ROCK Desktop (R3 Pro V2) | Pix4D | DroneDeploy |
|---|---|---|---|
| Primary data used for orthorectification | Fuses co‑aligned LiDAR point cloud and RGB photos. The LiDAR provides an accurate 3‑D surface, and the precise GNSS/IMU trajectory ensures each image is correctly positioned. Photos are rectified onto the LiDAR surface, reducing distortions. | Uses photogrammetry algorithms (feature matching, bundle adjustment) to reconstruct the scene from overlapping images. Optionally uses RTK/PPK data or import of camera positions to aid alignment but does not natively fuse LiDAR. | Similar to Pix4D; a cloud‑based platform that performs feature‑matching and bundle adjustment on images. Flight planning and automated capture ensure sufficient overlap, but orthorectification relies solely on image data. |
| Accuracy and corner capture | Rock’s R3 Pro V2 introduces a unique technique for precise corner capture; improved IMU/GNSS and LiDAR data lead to enhanced corner accuracy. The orthophoto inherits the centimetre‑level accuracy of the LiDAR. | High precision photogrammetry with adjustable parameters; accuracy depends on flight planning and the quality of the camera and GNSS. Does not inherently capture LiDAR‑level corner precision. | Designed for ease of use; provides moderate accuracy suitable for construction and agriculture. Accuracy depends on ground control points (GCPs) and RTK/PPK data; less suited for millimetre‑level corner capture. |
| Workflow | Process data locally in ROCK Desktop. Steps include trajectory computation, LiDAR colourization and orthomosaic generation. Orthophoto generation is integrated into LiDAR processing; no need for third‑party software. | Desktop (Pix4Dmapper) or cloud processing. Users upload images and optional CSV with camera positions. Pix4D performs photogrammetric alignment, dense matching and orthomosaic production. | Fully cloud‑based. Users plan flights and capture data using the DroneDeploy app; images are uploaded to the cloud for processing and orthomosaic generation. |
| Ease of use | Requires the R3 Pro V2 hardware and ROCK Desktop license. Workflow is tailored for survey‑grade LiDAR users; knowledge of LiDAR processing is beneficial. | Offers advanced controls and quality reports; more complex than DroneDeploy but provides flexibility. | Emphasises ease of use, automated flight planning and real‑time mapping; targeted at construction, agriculture and inspections. |
| Integration with LiDAR | Native integration; LiDAR and imagery are collected simultaneously with a co‑aligned camera. Orthophoto generation uses LiDAR geometry and accurate trajectory. | LiDAR integration requires separate workflow or external tools (e.g., merging LiDAR point clouds after processing). Not inherently built into the photogrammetry engine. | Does not integrate LiDAR; orthophotos are generated solely from imagery. |
| Resulting orthophoto | Seamless, distortion‑free orthophoto with LiDAR‑level georeferencing. Not limited by vegetation or low‑texture areas because LiDAR anchors the geometry. Orthophotos are exported as .tif files and can be displayed in ROCK Cloud. | High‑quality orthomosaics; accuracy depends on GCPs and flight planning. Susceptible to distortion in vegetated or low‑texture areas. | Quick orthomosaics optimized for progress monitoring; less accurate than Pix4D or ROCK Desktop. Ideal for rapid visualization and simple measurements. |
Processing time and scalability
The time to generate an orthophoto in ROCK Desktop is proportional to the number of photos. Generating multiple orthophotos from separate flights is not yet supported; each project produces its own orthophoto.
Benefits of ROCK Desktop Photogrammetry
- High‑accuracy orthophotos. Fusing LiDAR and RGB data provides centimetre‑level accuracy. The co‑aligned camera ensures near‑perfect overlap between the LiDAR intensity view and the RGB view, which reduces distortions and improves positional accuracy.
- Reliable corner detection and object location. The R3 Pro V2’s unique photogrammetry technique allows precise capture of building corners and small objects. Surveyors can delineate features without labour‑intensive ground control.
- Improved colorization and repeatability. Enhanced IMU, GNSS and processing algorithms result in better colorization and more consistent datasets.
- Unified LiDAR and photogrammetry workflow. Users collect LiDAR and imagery in a single flight. The workflow in ROCK Desktop automatically processes both data types, writes EXIF metadata and generates orthomosaics. This reduces manual steps and eliminates the need to export data to third‑party photogrammetry software.
- Flexibility for downstream use. The orthophoto output (.tif) can be viewed in ROCK Cloud alongside point clouds and surface models. It can be exported to CAD/GIS for planimetrics and mapping. For users who prefer other software, the georeferenced images and PPK_imageList.csv file can still be processed in Pix4D or other photogrammetry tools.
Conclusion
Rock Robotic’s ROCK Photogrammetry represent a significant shift in orthophoto generation. Traditional photogrammetry software like Pix4D and DroneDeploy rely on image matching algorithms and optional GNSS data to reconstruct scenes; accuracy depends on flight planning, overlap and ground control. ROCK Desktop uniquely fuses a co‑aligned RGB camera with a high‑accuracy LiDAR sensor and precise PPK trajectory. This combination allows the software to project each pixel onto a 3‑D LiDAR surface, producing orthophotos with LiDAR‑level georeferencing and improved accuracy. While Pix4D and DroneDeploy excel at scalable, cloud‑based photogrammetry, they do not inherently integrate LiDAR data. ROCK Desktop therefore offers a powerful solution for surveyors who require both high‑density point clouds and accurate orthophotos from the same flight.