TSPS MAR 2025 FINAL REV - Flipbook - Page 17
TECHNOLOGY
C1-PoiBnt Cloud Comparison
As shown in these two images generated in Carlson Point Cloud
Advanced from LAS 昀椀les are presented. The point cloud produced by DJI is denser than RIEGL’s, and the DJI LAS 昀椀le size
is three times larger. Additional 昀氀ights with RIEGL, with closer
昀氀ight lines, yielded similar results.
RIEGL-Single beam sensor
.LAS size 9,759,891 KB
DJI-Multi beam sensor
.LAS size 23,183,872 KB
Note: Although there were
no signi昀椀cant differences between the 昀椀rst and
second 昀氀ights with RIEGL
on the GCPs, DJI, despite
having a better average
elevation, showed high
differences in some GCPs.
The automatic report from
LidarMill, though showing
a smaller difference, is
comparable to the results
presented here, especially
on GCP 8.
Note for Images on Left: DJI, by capturing more returns, generates more
data, including noise points. The RIEGL processing is considerably more
accurate in de昀椀ning the natural terrain and in delineating break lines.
Note: In Topic A-1, the GCPs align very well with RIEGL, while some points
with DJI show errors greater than 0.4. In Exhibit 2, the XML exported by
RIEGL for topographies is excellent. In Exhibit 3, it is necessary to consider
whether adjustments should be made to obtain a denser point cloud with
RIEGL or if, due to its greater precision compared to DJI, the current cloud
is suf昀椀cient, although it appears less dense visually.
2. TERRESTRIAL SCAN DATA: RTC 360
A2- Data collection (terrestrial scan), which is not adequately covered
by aerial Drone Lidar
The terrestrial data was strategically distributed to cover the
entire area of the structure, ensuring that we cover all areas that
the drone data cannot clearly reach and considering the maximum distance of the RTC equipment to link the setup points,
which is 50 feet, and ensuring data consistency to achieve a
昀椀nal bundle with proper calibration.
March 2025 THE TEXAS SURVEYOR
15
