Agisoft Metashape allows users to create highly detailed 3D models by combining both aerial and terrestrial (ground-level) photos. This hybrid approach is especially useful when modeling buildings, archaeological sites, industrial areas, or any structure where drone and DSLR images offer complementary perspectives. In this article, you’ll learn how to integrate both types of imagery into a single, seamless photogrammetry project.
Why Combine Aerial and Terrestrial Images?
While drones offer an excellent top-down view, they often miss important side and under-structure details. Ground-based images fill in these gaps by capturing vertical facades, interiors, and occluded areas with higher detail.
- Improved coverage: Minimize data voids and occlusions
- Increased accuracy: High-resolution DSLR cameras enhance detail in critical zones
- More complete 3D reconstruction: Ideal for heritage documentation, architecture, or urban modeling
1. Plan Your Capture Strategy
Start with clear planning. Capture your aerial images with a drone using high front and side overlap (typically 80%/70%). Then, use a handheld or tripod-mounted camera to capture terrestrial images, especially around facades, entrances, statues, or interiors. Ensure adequate lighting and consistent camera orientation.
- Use manual settings to ensure consistent exposure
- Capture enough overlap between aerial and ground images
- Consider using Ground Control Points (GCPs) or markers visible in both datasets
2. Organize and Import Images
Place all aerial and terrestrial images into separate folders for clarity. In Metashape, create a new project and add both sets of images into the same chunk. If needed, you can also import them as separate chunks and merge them later (we’ll cover this below).
In the photo pane, tag or label each group for easy reference. This helps during troubleshooting or alignment refinement.
3. Calibrate Camera Groups Separately
Drone images and DSLR images typically have different sensor sizes, focal lengths, and distortions. Make sure Metashape recognizes the different camera groups and applies calibration accordingly. You can do this by checking the “Camera Calibration” pane and assigning parameters per camera group if necessary.
For better accuracy, consider running “Optimize Cameras” after alignment to adjust internal parameters using GCPs or tie points.
4. Align Photos
Run the “Align Photos” step for the entire dataset. Select “High” accuracy and enable the “Generic preselection” and “Reference preselection” options. This will generate tie points and establish the relative positions of all cameras in one coordinate system.
In the sparse cloud, verify that both aerial and terrestrial camera positions are aligned correctly and overlap visually. Remove any poorly aligned images using the “Show Errors” tool.
5. (Optional) Work with Multiple Chunks
If you’ve aligned the aerial and ground images in separate chunks, you can merge them after alignment. Use the “Merge Chunks” tool and select “Cameras” and “Tie Points” options to integrate the datasets. This is useful if images were processed in separate sessions or by different teams.
6. Build the Dense Cloud and Mesh
With everything aligned, proceed to “Build Dense Cloud.” Use “High” or “Medium” quality depending on your system’s hardware. Filter mode can be left as “Mild” for mixed imagery sets. Then build a mesh from the dense cloud for 3D modeling.
In cases where certain parts of the mesh appear noisy, use depth maps or apply selective masking on problematic images before dense cloud generation.
7. Texture and Export
After building the mesh, create a texture using “Mosaic” blending mode. Make sure the UV mapping is set to “Generic” for general-purpose visualization. You can now export the model in formats such as OBJ, FBX, or GLB for use in external applications like Blender or Unity.
8. Accuracy Tips and Best Practices
- Use coded or manually placed markers visible in both aerial and terrestrial images
- Calibrate each camera separately and optimize alignment
- Mask sky or irrelevant backgrounds in ground images to improve alignment
- Use the “Gradual Selection” tool to clean sparse cloud noise before dense reconstruction
Conclusion
Combining terrestrial and aerial photos in Agisoft Metashape enables you to build highly complete and accurate 3D models. Whether you’re documenting architecture, inspecting infrastructure, or preserving cultural heritage, this hybrid workflow delivers the best of both worlds. With careful planning, separate calibration, and thorough alignment, you can achieve professional-grade results that are ready for analysis, visualization, or presentation.