πΊοΈ3.3 Get3D Mapper
Get3D software is a powerful photogrammetry 3D model reconstruction software, supporting processing large scale image AT and model reconstruction. It is able to process millions of image 3D modeling.
Last updated
Get3D software is a powerful photogrammetry 3D model reconstruction software, supporting processing large scale image AT and model reconstruction. It is able to process millions of image 3D modeling.
Last updated
Get3DMapper is the main user interface, the software project is a tree-like organisational structure, which contains one and multiple blocks under the project; different reconstruction frameworks can be created under the blocks according to the requirements; and multiple different types of product results can be submitted under the reconstruction frameworks.
After opening Get3D Mapper software, click "New Project" or use the shortcut key "Ctrl+N". Open the New Project page. On the New Project page, you can set the project name and project location. If cluster computing is required, set the project path to the network path.
Click Open Project, select the *.m3d project file, and click the "Open" button. You can also open the project by clicking "Project Name.m3d" in the "Recent Projects" list.
Click the "Project" menu, click the "Save Project" button or use the shortcut key "Ctrl+S" to save the project information.
Project processing is complete, click the "Project" menu, use the "Clean up Project" function to clean up the completed project, the project has a task to run, you can not perform this operation, does not contain the product results generated by the software and intermediate results.
β γClean Medium Cache
Clear the cache files in the project with one click, including project cache (preview), block cache (feature points, matching, leveling), and reconstruction cache (dense matching point cloud).
β‘γClean Reconstruction Tile Cache
Clear the cache of reconstructed (densely matched point clouds).
β’γSelective Deletion
Filter by Project Cache, Block Cache, Rebuild Cache, and check Cache File Directory Delete.
Click the "Close Project" button in the "Project Menu" of the software, and select "Yes" in the pop-up dialog box to close the currently open project.
Click the "End" button in the "Project Menu" of the software or use the shortcut key "Ctrl+W" and select "Yes" to close the main interface of the software.
Add Images
Get3D Mapper supports importing aerial photos and ground photos, you can organise your data according to the "Get3D File Organisation Specification", and choose to import as needed according to the actual scene.
After the data is imported, the image needs to be positioned further to restore the pose of the image. Get3D Mapper supports positioning types such as exif (Exchangeable Image File Format) positioning and using pose file positioning.
a. Confirm photo Exif information
b. Adding photos
Select the root directory of the file to import all the photos under the folder.
c. Read Exif location
Check the Exif information column, select the "exist" photo group, right click to open the menu bar, click "Read Exif" button.
d. Query
Ask if the prompt to read Exif information is slow and whether to continue to prevent misuse.
e. Progress bar "Reading photo location".
Progress bar for reading Exif information.
f. Positioning completion
After the positioning is completed, "0/2250" is switched to "2250/2250", i.e. each photo in the photo group has positioning information corresponding to it.
g. Adding photographs
h. Automatic matching of photo groups and Pose files
a) Automatic Matching
Pose files are stored in the way shown in the following figure, Pose file name contains keywords such as POS; GPS; RTK; PPK, etc. After importing photos, the photo group will automatically correspond to the Pose file.
Tip: Pose file does not contain (POS; GPS; RTK; PPK) keywords, you can set the keywords manually, select the photo group, right-click and use the menu bar, automatically search by keywords.
b) Manual correspondence
Batch select multiple groups of photos by right-clicking to open Select Pose File in the menu bar.
i. Selection of the Pose file
j. Define the Pose format template
Set the spatial reference of the Pose, the separator, and the actual meaning of each column ("photo name, longitude, latitude, elevation" or "photo name, east coordinate, north coordinate, elevation").
k. Apply to photo groups
After defining the template, click "Apply to Photo Group" to select the photo group.
l. Positioning completion
After the positioning is completed, "0/2250" is switched to "2250/2250", i.e., each photo in the photo group has positioning information corresponding to it.
Add Video Frames
Get3D Mapper supports adding videos to get the corresponding video frames and add them to the block image. In the data management interface, click the "Add Video Frames" button, set the input and output file paths, and set the start and end of the extracted video frames as well as the interval time, then you can extract the imported video frames with one click.
Add Mobile Laser Point Cloud
Get3D Mapper supports the addition of laser point clouds and track line data collected by mobile laser equipment. The software supports laser point cloud modelling alone and laser point cloud fusion modelling with photos.
Support vehicle-mounted, airborne, backpack, and handheld laser point clouds. Laser point cloud supported formats: *.las, *.ptx, *.pts, *.e57. Trajectory line supported formats: *.txt, *csv.
1) Open the Add Moving Laser Point Cloud interface
2) Setting up the laser point cloud spatial reference
3) Selection of moving laser point cloud data
4) Importing trajectory lines
5) Define the trajectory line
Set the spatial reference of the track line, the separator, and the actual meaning corresponding to each column ("GPS time, longitude, latitude, elevation" or "GPS time, east coordinate, north coordinate, elevation").
6) Enquiry
Whether you need to check for data matching.
7) Check the matching results
Laser point cloud check contains laser point position and track line position matching; laser point cloud time and track line time matching. And display the matching status.
8) Data management interface display
9) Select the point cloud in the list
Add Static Laser Point Cloud
Get3D Mapper supports adding laser point clouds and center data collected by a standing laser machine. The software can support laser point cloud modeling alone and laser point cloud and photo fusion modeling. The following formats are supported: *.las, *.ptx, *.pts, and *.e57. The station center supports manual input or import of *.txt format files.
Add Control Points
Add control points to the data management view, as a block split reference, or as an imported photo data integrity reference.
a. Open Add Control Point
b. Selection of control point documentation
c. Define the control point file format
Set the spatial reference and delimiter. Define the control point file as "point name, longitude, latitude, elevation" or "point name, east coordinate, north coordinate, elevation".
The following is an introduction to the interface functions:
d. Successful import of control points
e. Clearing control points
The added control point reference data can be clicked on and "Clear Control Points" cleared in the view.
Add KML
KMLοΌ which stands for Keyhole Markup Language and was originally developed by Keyhole Corporation, is a coding specification based on XML syntax and formatting for describing and preserving geographic information such as points, lines, images, polygons, and models.
a. Select the background KML file
Add a range line to the Data Management view as a reference to the data integrity of the imported photos.
Range line reference data can be imported after tilt or other multi-source data is imported in the data management interface.
b. Select the background KML file
c. Delete KML line
The imported range line data can be cleared by clicking "Delete Range Line".
After adding photos and pose data, in the data management view each photo is expanded according to the corresponding position (x y z), and editing operations such as selecting, truncating, framing, and deleting can be performed.
The photo visualization interface in the data management interface, in the photo group list view, you can rename, block and other operations on the positioned photo group data.
Perform save operation for edited Aerial Triangulation. For example, delete photos, connection points, and control points to save the edited Aerial Triangulation results.
Aerial Triangulation results can be exported to do Aerial Triangulation results backup or re-imported into other projects to use. Reconstruction Master supports the import of Aerial Triangulation formats: *.xml, *.xbin.
Get3D Mapper supports the export of internally generated data such as connection points, marking information for stab points, etc. in the format of *.xml, *.xbin.
If the original photo paths have changed, to ensure that the Aerial Triangulation and Reconstruction tasks read these photos properly, you need to update the paths of the photos in the block.
Open the block photo group path modification screen.
Click "Check Update" at the bottom of the screen to check before making changes.
Red: The path of the photos recorded in the block is abnormal, and there is no corresponding photo under the path.
Green: The corresponding photos can be found in all the photo paths recorded in the block.
According to the original path, find the sibling folder of the new path and click to select it. The photo storage structure has not changed, and other paths can be replaced automatically.
For damaged or missing images, you can delete them by clicking "Clear missing photos" below.
Aerial Triangulation consists of two parts, the free net and the control net.
Relative Orientation: Relative Orientation, i.e. free net air three, refers to the work of restoring or determining the relative relationship of the image pairs at the time of photography, i.e. solving the work of the relative orientation elements of the stereo image pairs. It is through the observation of the image of the object in different viewpoints of the position of the relationship to determine the internal and external parameters of the camera, the relative orientation of the stereo image pairs is to restore the photography of the two neighbouring image beams of the interrelationship between the two, so that the same name of the light pairs to the intersection of the light pairs.
Absolute orientation: absolute orientation that control network air three, is through the measurement of the ground control points, the image and the ground coordinate system for contact. It uses the known ground control points to translate, rotate and scale the relatively orientated three-dimensional geometrical model to incorporate it into the ground photogrammetric coordinate system, and to improve the absolute accuracy of air three through the control points.
After the photos are imported into the software, the first step needs to be a free net to recover the internal and external parameters of each photo; the second step is to make a control net to improve the absolute accuracy of the Aerial Triangulation.
Aerial Triangulation parameter setting
The preset options, which contain more settings, are shown below:
3D View
The content displayed in the Aerial Triangulation 3D view includes information such as photo attitude, connection points, control points, laser point cloud, and laser trajectory lines.
Aerial Triangulation 3D Views
Aerial Triangulation Report
The Aerial Triangulation report mainly includes the following parts: project overview, camera calibration, camera information, connection point information, and control points.
Project overview: It contains the project where Aerial Triangulation is located, the running time, the number of photos, the number of control points, the number of photos entering the network, the rate of entering the network, the number of connecting points, the average resolution of connecting points, and the re-projection error of connecting points.
Camera calibration: A detailed list of information on the internal and external parameters and distortion coefficients of the camera set, which are used to assess the calibration accuracy of the camera.
Camera information: Schematic of the offset distance of the camera position from the original POS data, schematic of the re-projection error for each photo, the number of connection points observed by the camera, and the scene coverage. Corresponding connection points, re-projection error for each photo.
Connection point information: number of observation times of connection points, re-projection error of connection points,
Control points: coordinates and error information of each image control point.
If the product generated only by inputting the photo positioning data does not meet the absolute accuracy requirement, it is necessary to add control points at the end of the free net Aerial Triangulation to get the control net result and improve the absolute accuracy of the product.
Some of the common shortcuts for Mark GCPs is shown:
Click the "Edit Control Point" button to enter the Edit Control Point screen.
Setting up an Aerial Triangulation reference.
Import Control Points: Click "Import Control Points" button to enter the interface of importing control points, select the spatial reference corresponding to the control point file, select the separator, set the name, X, Y, Z, and click "Confirm" to import the control points.
Mark GCPs: Selecting a control point in the Control Points list or double-clicking on a selected control point in the block view enters the editing state for that control point. The "Photo Preview" screen on the right shows the visual image of the control point.
Switch to "Photo View", judge the actual location of the point according to the mark of the point, "Shift key + left mouse button" to do the stabbing point operation.
Select by mouse or "A/W/S/D" to switch to the next photo stop, "Shift + left mouse button" to do the stabbing operation. Repeat this step, 20-30 photos are recommended for five-lens flight data, and 10-15 photos are recommended for single-lens flight data.
Create a new 3D reconstruction spatial frame on Aerial Triangulation (set up spatial references, area of interest ranges, tile divisions, reconstruction memory estimates, and more settings) under which one or more products can be submitted.
(1) Spatial reference
The spatial reference setting for the rebuild frame uses the ENU coordinate system (Local East-North-Up system) by default, which can be set to either an existing coordinate system or a user-defined coordinate system. Setting the same spatial frame ensures that the later rebuilt tile boundaries pick up exactly where the previous tile boundaries left off.
(2) Range
Adaptation to Connection Points: The area of interest ranges to the largest outer enclosing box that contains all connection points.
Adaptation camera: the area of interest ranges to the largest outer enclosing box containing all cameras.
(3) Tile division
Division mode: planar grid division, 3D grid division, no division.
Tile length: set the tile length. There is a proper tile edge length by default (related to null3 resolution, related to GSD), user can customise the tile edge length.
Buffer Radius: set the overlap length of neighbouring tiles, when using tile rebuild, in order to avoid gaps between tiles, set the width of overlap between tiles.
Origin XYZ: set the coordinates of the origin of the slice.
Real-time preview: after modifying the parameters, display the updated result in the 3D view in real time.
Ignore empty blocks: tiles without connection points are not rebuilt.
(4) Space frame saving and loading
Save space frame: save the reconstruction space frame settings, including space reference, tile division origin, tile size, buffer radius.
Load space frame: load the reconstruction space frame settings.
(5) Memory estimation
Single tile using the maximum memory value prediction, tile reconstruction requires the size of the computer memory related to the tile size (the larger the tile the larger the computer hardware memory required), according to the computer memory to determine the size of the tile division, the prediction of too much memory will lead to the failure of the tile reconstruction.
For example, if you have 64GB of computer memory, you can adjust the tile length, set the tile estimated memory to 30GB, and actually divide the largest tile.
(6) Copying reconstruction parameters
Can copy other reconstruction parameters under the same project contains space reference, tile division origin, tile size, buffer radius.
More Reconstruction Settings interface allows you to set the Point Cloud Generation Strategy, Geometry Generation Strategy, and Texture Mapping Mode in the reconstruction step. Different preset reconstruction modes can be set in More Reconstruction Settings for different modelling scenarios:
β Point cloud generation
Mainly image dense matching strategy, photo selection strategy, matching strategy, matching accuracy.
β‘ Geometric Generation
White model construction:
3D conformal mesh: suitable for most model reconstruction cases.
2D mesh: suitable for downward looking image modelling.
Model simplification strength:
Weak: low simplification effort, denser triangular mesh, more (valid and invalid) geometric details, larger data volume.
Medium: moderate strength of simplification, with a balance of data size and (valid and invalid) geometric details.
Strong: strong simplification, sparser triangular mesh, largely filtered (valid, invalid) geometric details, smaller data size.
Model OptimisationοΌ
When ticked, the 3D model contains more details.
Topological edge optimisationοΌ
Topology optimisation of the geometry between tiles can effectively avoid gaps between tiles.
β’ Texture mapping modes
Balance mode: default texture mode, texture clarity and error projection.
Correctness Priority: Applicable to urban modelling scenes, improve the texture misprojection problem for moving targets, fine objects, trees.
Clarity Priority: Applicable to close-up modelling scenes, select the image with the highest clarity among multi-scale images.
β£ 3D View Function
Layer show/hide: Show/hide control for area of interest, tile, camera, connection point, point cloud, control point, constraint layer.
Reset View: Reset the 3D view.
Camera increase +: increase the size of the camera in the 3D view.
Camera Decrease -: increase the size of the camera in the 3D view.
Edit Area of Interest: Manually adjust the reconstruction range in the 3D view, select and then determine the area of interest, i.e. the modelling area, by extruding and stretching the hexahedral enclosure box.
Import KML: Import KML range for reconstruction range constraints, output 2D and 3D model results can be output by range line shape.
Export Tile Boundary KML: For the selected tile, export tile boundary as KML file.
Get3D Mapper can output 3D model products in OSGB and OBJ formats, 3D point cloud products in LAS, TXT, ASC and E57 formats, digital orthophoto products in TIFF and JPG formats, and digital surface model products in TIFF and ASC formats.
Several different types of products can be submitted under the same reconstruction framework.
β 3D model products
After the Aerial Triangulation processing is completed and the reconstruction settings are finished, you can submit the 3D model products in OSGB and OBJ formats for generation.
Select the output format, texture source, texture quality, maximum texture size, and whether to apply a tonal style.
β‘ 3D point cloud products
After generating a 3D model, the model is sampled to produce a 3D point cloud product.
The software interface allows you to set the point cloud product output format, point cloud source, and resolution.
After generating the 3D model, the model is sampled to generate DOM and DSM products.
The software interface allows you to set the resolution, split image size, DOM product format, and DSM product format.
Water surface fluctuation, and the existence of reflection and refraction and other phenomena make different photos taken by different photographs of the same water surface photographs expressed on the inconsistent, the lack of sufficiently stable feature points of the same name to build the water surface structure. In order to solve this problem, the software provides manually drawn constraints to assist in constructing the water surface structure, effectively solving the problem of large holes in the water surface on the model.
The list of geometric constraints contains the imported constraint ID, constraint name, constraint type, height value, colour, commit status, and import time.
The Aerial Triangulation view interface includes a "Layer" control.
β Drawing constraints
There are two ways to draw the water shoreline, drawing on the empty three results, using Get3D Viewer software to open the 3D model refinement drawing.
Aerial Triangulation results drawn range line precision is not high, do not need to secondary update processing, but the final output of the water surface effect is poor.
The range line drawn by the 3D model has high precision, requires secondary update processing, and the final model has good results.
1) Aerial Triangulation results drawing
Click the "Draw Constraints" button, left-click on the Aerial Triangulation views of the water surface shape, adjust the view angle to adjust the height of the water surface constraints, set the constraint type (water, island).
2) Three-dimensional model drawing
Accurately draw the water surface range, three-dimensional reconstruction results after the completion of the production, click on the "View Product", linkage DasViewer to open the three-dimensional model, use the constraints drawing function, drawing along the waterfront, double-click to close after completion.
Draw the island.
Use the colour picker to select the appropriate water colour on the model. The water surface colour selection includes three colour setting methods Photo Colour Pickup, Swatch Colour Selection, and Manual Setting (Hue Saturation, Luminance, Saturation, R, G, B) values.
Export the water surface constraint file in xml format.
Import the Reconstruction Master software in the Geometric Constraints screen.
β‘Checking Constraints
Preview the imported constraints to check whether the position, height and colour are reasonable. If there is no problem, click "Submit Constraints" button.
β’Submit update
Constraints drawn and imported after 3D reconstruction, after submitting the constraints, submit the 3D model product and the constraints will take effect when the model is generated.
Constraints drawn and imported before 3D reconstruction.After submitting the constraints (submitting the constraints deletes some of the intermediate results of the constraints on the gland tile), you need to select the tile with the status of "To be updated" and the update type of "Surface constraint" in the product list.
Once selected, click the "Update" button.
View the results of the water surface constraints.
The tool contains 3 functions: merge DOM/DSM, clean up and rebuild intermediate results, and cluster diagnostics, which can be opened from "Tools" in the menu bar.
Merges the generated DOM/DSM split images into a whole image. The specific use operation is as follows.
β Select the corresponding "Block", "Reconstruction" and "Product".
β‘ Check the "DOM" and "DSM" of the products to be merged and set the output path;
β’ Sampling resolution can be maintained and custom sampling;
β£ Import the merged range or set the range threshold, and cut according to the range;
β€ Click "Merge" button to start merging DOM/DSM.
Clean up and rebuild the intermediate results of the tool, you can clean up the specified tile under the specified type of intermediate files, to release the storage of disk space. The specific operation is as follows.
β Select the corresponding "Block" or "Reconstruction";
β‘ Select the tile where the intermediate results need to be cleared;
β’ Select the corresponding "Block" or "Reconstruction".
β£ Select the tiles to be cleaned up;
β€ Click "Clean" button to clean up the selected intermediate file types under the selected tile.
Note: This function cannot be used when there are running tasks in the project.
Cluster environment diagnosis includes three aspects of "hardware environment", "network environment" and "network throughput".
You can open the Cluster Environment Addendum tool to get a summary of system information; check whether each engine has access to the project directory, original image directory, and task directory; and estimate the maximum number of supported engines based on the current network bandwidth and network throughput.
If you encounter technical problems, you may need to provide information about the cluster environment and seek technical support.
Get3D Mapper generates detailed Aerial Triangulation Reports and Reconstruction Reports, allowing users to learn more about the details of aerial triangulation and reconstruction.
For projects that have completed aerial triangulation, an Aerial Triangulation report can be generated by clicking on View AT Report in the block view.
For projects with large image data, the process of generating this report takes some time.
The AT report contains an overview of the project, camera corrections, camera information, and tie point information. For projects that include control points information, the aerial triangulation report also contains control point information.
The project overview provides information on aerial triangulation time, mean ground resolution of photos, ground coverage, and the proportions of the photos in the nets.
The correction of camera information contains the camera distortion parameters before and after calibration.
The camera information can be viewed as a legend with information such as the offset distance of the camera from the original POS data, the reprojection error for each photo, and the number of connection points observed by the camera, the scene coverage, and so on.
The tie point information can be visualized as a graphical representation of the number of observations and the reprojection error of the connection point.
For the project that includes control points, the aerial triangulation report also contains control point information. In the control point information, you can visualize the error distribution of control points and check points.
In the product interface, you can click on View Product Report.
The reconstruction report contains Project Overview Information and Tile Information.
The project overview contains information about the coordinate system and origin of the modeled tiles, the size of the modeling area, and the time spent on the modeling.
Tile information allows you to see how long it took to generate each tile during the modeling process.
