v7.1.3
Added rapid rough registration function for aerial triangulation of images without location metadata and laser point cloud.
Optimized aerial triangulation algorithm to improve stability and accuracy for images with high geolocation accuracy captured by DJI lenses.
Optimized laser point cloud modeling process and adjusted reconstruction parameters.
Optimized software interfaces and icons.
1 Added Rapid Rough Registration Function For Aerial Triangulation Of Images Without Location Metadata And Laser Point Cloud
There are deviations in position and scale between the aerial triangulation of images without position metadata and the laser point cloud. You can use the newly added Coarse Registration function in the GCP marking interface to address this and align the aerial triangulation tie points to the laser point cloud.
(1) Result Comparison
a. Before Alignment
b. After Alignment
(2) Operation Instructions
For aerial triangulation blocks of images without position metadata that contain laser point cloud data:
a. Click the Edit GCP button to access the GCP marking interface;
b. Click 3D View → Registration;
c. Access the Coarse Registration interface;
d. The left side is the laser point cloud view, and the right side is the aerotriangulation tie point view;
e. First pick a point in the left view, then pick a homologous point in the right view. Select at least three pairs of homologous points;
f. Click the Preview button to view the alignment result;
g. Click the OK button to apply the alignment result.
2 Optimized the aerial triangulation algorithm to improve the stability and accuracy of aerial triangulation for high-precision position images captured by DJI lenses
(1) Model Comparison (Left: Aerial Image Modeling; Right: Aerial + Low-Altitude Supplementary Image Fusion Modeling)
(2) Operation Instructions
When more than 50% of the photos in the Block are identified as having high-precision position information, the High Geolocation Accuracy mode will be displayed and checked by default when submitting the aerial triangulation task. This mode significantly improves aerial triangulation stability and accuracy.
(3) Method to Determine Whether Photos Have High-Precision Position Information (Only fixed solution positioning with an RTKFlag of 50 provides centimeter-level high-precision position information.)
Table 1 RTK Accuracy Levels
a. Method 1: MRK file – the last line displays the RTK position accuracy level.
b. Method 2: Open the photo in text mode (garbled characters are normal), then press Ctrl + F to search for the "RTKFlag" field.
(4) Aerial triangulation Results
a. Aerial images and close-range images are processed separately; select them and then merge the blocks.
b. Process aerial images and close-range images together in a single block for aerotriangulation.
3 Optimized the laser point cloud modeling process and modified reconstruction parameters
(1)Added the non-simplification modeling mode
a. Result Comparison
b. Operation Instructions
In the Reconstruction Setting, select the Disable option for the Mesh Simplification parameter to enable the non-simplification modeling mode.
(2) Point Cloud Modeling with Image Texturing
a. Use point cloud data for modeling and image data for texture mapping.
b. Operation Instructions
In the Reconstruction Setting interface, under Point Cloud Generation, check only the Laser Point Cloud option to use only laser point clouds for modeling;
In the Product Setting interface, check the Oblique Image option for the Texture Source parameter to use images for texture mapping.
4 Optimized the software interface and icons
(1) Data Prepare Interface
(2) Submit AT Interface
(3) Point Cloud Tab Interface
(4) Reconstruction Setting Interface
(5) Product Interface
Last updated