
Megadron Blog


Unlock more advanced mapping and modeling capabilities with the DJI Terra Standard to Flagship Upgrade (Offline). This electronic license upgrade is a great choice for users who want to take their DJI Terra workflow to the next level and gain access to more powerful tools for 3D reconstruction, LiDAR processing, and professional spatial analysis — all in a convenient offline authorization format.

DJI Terra is an all-in-one 3D reconstruction and mapping software suite that combines advanced photogrammetry with LiDAR data support. The Flagship version expands the software’s capabilities, making it a smart upgrade for users working on more demanding projects, larger environments, and more detailed digital models.
One of the biggest advantages of the Flagship version is support for 3D Gaussian Splatting. This technology allows DJI Terra to generate highly realistic 3D models from photos, making it easier to recreate buildings, monuments, vegetation, reflective materials, and other visually complex structures. Compared to traditional oblique photogrammetry, it handles transparent elements, reflections, and fine details more effectively while also helping create cleaner, less distorted orthophotos.

DJI Terra is built for speed and efficiency. It can process around 500 photos per hour, and for single-block modeling, reconstruction can take only about 30 to 40 minutes. In projects with more than 10,000 images, the upgraded technology can operate up to twice as fast as the earlier Terra Mesh engine, helping you complete large jobs more smoothly and with less waiting.

The Flagship upgrade is ideal for users working with advanced urban, industrial, and infrastructure scenarios. DJI Terra can process approximately 300 to 400 photos per 1 GB of available RAM, and a single reconstruction task can include up to 30,000 images. That means you can build detailed models of entire districts, industrial parks, or utility networks with impressive flexibility.

Despite its advanced tools, DJI Terra remains accessible. For smooth operation, a graphics card with 4 GB of VRAM and 32 GB of RAM is enough. This makes the software easier to run without needing an ultra-expensive workstation, which is great for enthusiasts and advanced users who want powerful results on more moderate hardware.

Finished reconstruction results can be exported with a single click to widely used formats such as PLY and B3DM (3D Tiles). This makes it easy to integrate DJI Terra output with external platforms used in surveying, urban planning, GIS workflows, game development, and film production.

DJI Terra can generate high-quality 3D models, point clouds, orthomosaics, and DSM surface models from photo data. You can also adjust 2D map resolution and 3D mesh density to match your project needs. When paired with DJI drones equipped with RTK, the software can produce maps with accuracy meeting 1:500 mapping standards without requiring ground control points.

DJI Terra is fast enough to generate results directly in the field for smaller projects. For larger 3D reconstruction jobs, it can process up to 30,000 images in three days on a single workstation, or in as little as 21 hours using cluster technology. For 2D reconstruction, 5,000 photos can be processed in under 2 hours, or around 80 minutes in cluster mode.

The Flagship version supports advanced integration of LiDAR data and visible-light imagery. A single flight with the Zenmuse L2 can produce precise point clouds, 2D maps, and 3D mesh models at the same time. This creates more complete and detailed reconstructions, even for complex objects such as bridges. With RTK and PPK support, centimeter-level accuracy is possible without ground control points. DJI Terra can also automatically create flight routes based on local 3D models or point clouds, including data from external sources.

DJI Terra can process approximately 1 GB of point cloud data per 1 GB of available RAM, allowing smooth handling of demanding LiDAR projects. In practice, a computer with 32 GB RAM can process up to 120 GB of data, covering an area of around 10 km². For city-scale datasets, systems with 128 GB RAM can be used for even greater performance.

DJI Terra supports data from the DJI Mavic 3 Multispectral and Phantom 4 Multispectral, enabling the creation of vegetation index maps such as NDVI and NDRE. It can also generate reflectance maps and spraying routes for DJI Agras drones, including variable-rate operations. Simply choose the crop type, and the software can generate optimized 2D maps to support smarter agricultural decisions.

DJI Terra is ready for distributed environments, making it possible to process huge volumes of data beyond the limits of a single workstation. In a setup with five working units, up to 30,000 images can be processed in about 21 hours. This gives you the flexibility to scale performance depending on project size.

DJI Terra works with a wide range of DJI Enterprise hardware, including DJI Matrice 4E, DJI Matrice 400, Zenmuse P1, Zenmuse L2, and the D-RTK 3 Station. Together, they form a complete workflow for photogrammetry and advanced LiDAR mapping.

With just one click, you can transfer a model from DJI Terra directly to DJI Modify, DJI’s mesh and point cloud editing software. There you gain access to powerful functions such as smoothing, denoising, sampling, terrain classification, and export to GIS and CAD applications. Combined with DJI drones, Terra and Modify create a streamlined workflow from data capture to editing and sharing.

DJI Terra also includes support for local and cloud-based PPK processing, control and check point marking, more than 8,500 coordinate systems, PRJ files, and seven-parameter transformations. Users can also perform measurements, add annotations, and export data in formats compatible with local mapping standards.

| Feature / Version | Agriculture | Standard | Flagship | Cluster | Education |
|---|---|---|---|---|---|
| Agricultural applications | ✔ | ✔ | ✔ | ✔ | ✔ |
| 2D mapping (farmland/orchards) | ✔ | ✔ | ✔ | ✔ | ✔ |
| 2D mapping (urban areas) | — | ✔ | ✔ | ✔ | ✔ |
| 2D multispectral reconstruction | — | ✔ | ✔ | ✔ | ✔ |
| 3D mesh reconstruction | — | ✔ | ✔ | ✔ | ✔ |
| LiDAR point cloud reconstruction | — | ✔ | ✔ | ✔ | ✔ |
| Terrain Result | — | ✔ | ✔ | ✔ | ✔ |
| Gaussian Splatting technology | — | — | ✔ | ✔ | ✔ |
| Energy sector applications | — | — | ✔ | ✔ | ✔ |
| Cluster computing | — | — | — | ✔ | — |
| Max. linked devices | 3 | 1 | 1 | 1 device, 3 nodes | 10 |
| Authorization mode | Online | Online/Offline | Online/Offline | Offline | Online |
| Validity period | 1 year | 1 year / perpetual | 1 year / perpetual | Perpetual | Perpetual |