DJI Terra Update Targets LiDAR Errors That Waste Survey Time
For once, enterprise-grade software for mapping doesn’t usually focus on eliminating drudgery, but that’s the area where most efficiency gains happen. DJI’s current upgrade to their Terra software, DJI Terra 5.2.5, is one such case. Rather than offering any cool feature like a fancy reconstruction mode, its key focus lies in addressing an issue familiar to many a survey team the disconnect between the field side and the LiDAR processing in the office. The most important update of the version is that it now automatically retrieves base station data during the Cloud PPK processing workflow, eliminating a step where users had to manually import files before doing any reconstruction. In LiDAR projects, accuracy matters greatly, as it affects whether one can trust measurements made afterward.
That’s because Terra’s become an integral part of a much bigger workflow, used in construction sites, corridor inspections, utilities, mines, and law enforcement. In fact, DJI’s software suite had already begun to develop toward greater integration with capture hardware, especially after releasing point cloud reconstruction capabilities for free with the Zenmuse L3 camera in version 5.1. At the same time, Terra’s distributed processing abilities for multiple computers on a network had improved to tackle the ever-growing volumes of data collected on large infrastructure projects. For survey departments working on the reconstruction of rail alignment, power line corridors, or extensive terrain models, the bottleneck has moved away from flying the mission itself toward post-processing.
This update also improves the algorithm for Multiple Time Around, helping to improve the quality of long-distance scanning while reducing the necessary processing time. Long-range LiDAR tasks are always especially prone to slight deterioration due to sparse points, difficult geometry, or vegetation-rich environments. As the developers have been leading up to this since the lidar functions appeared in version 3.9.0 in October 2024, followed by the addition of visible light integration, direct support for newer aerial platforms for PPK, and the advanced Gaussian Splatting technique. Other improvements might be less immediately apparent.
DJI Terra 5.2.5 fixes an issue causing incorrect calculation of coordinate values in the CSV export of grid-of-points and TIN results after applying seven-parameter transforms in visible-light tasks. Also addressed are incomplete fields in LiDAR quality reports, partial colorization of point clouds when Smooth Point Cloud was selected, and a visualization bug that would render the base-station center point height as zero in certain cases. On the surface, none of those issues seem too big individually. Collectively, however, they can affect critical artifacts sent further along to GIS, engineering, and asset management teams.
And that brings us to another interesting aspect of DJI’s current development strategy. In the current context, software for mapping isn’t simply something you use to complete a project and move onto the next one. Third-party applications like LP360’s L1/L2 processing chain make use of Terra for generating trajectories and initial point cloud data, after which additional processing happens. From that standpoint, any improvements in automating PPK input and reporting contribute immensely to smoothing the process of passing artifacts through.
In other words, DJI Terra is becoming less focused on flashy functionality and more on removing points of potential failure in enterprise LiDAR projects. That ranges from correcting positions and making point clouds look right to processing them efficiently via a distributed approach. This is easier to understand for users more efficient operation, less need to fix manual mistakes, and faster completion.
