Bundle block adjust UAV or UAS data

Author a mosaic dataset

The first step to bundle block adjust your unmanned aerial vehicle (UAV) or unmanned aircraft system (UAS) imagery is to create a mosaic dataset and then load your data into it.

UAV data is stored as JPEG images, with the orientation information stored in a JPEG Exif header, or as a GPS table that is distributed with your data. The GPS table should include metadata such as latitude, longitude, and altitude, and can optionally include Omega, Phi, and Kappa.

1. Use the Create Mosaic Dataset tool to create an empty mosaic dataset container.
   1. The Coordinate System should be set to the UTM zone of the input images.
   2. Click OK to run the tool.
2. Open the Add Rasters to Mosaic Dataset tool.
3. Choose UAV/UAS for Raster Type.
4. Set the input data as the folder location of the UAV or UAS data.
5. Click the Raster Type Properties button to open, view, and edit the properties for the raster type.
6. If you have GPS or orientation file, click the Auxiliary Inputs tab, and specify the file and location.
7. Click the Properties tab.
8. In the Orthorectification using elevation section, click the DEM option and specify your digital elevation model (DEM) input.
   • If no DEM is available, and your study area is relatively flat, you can check the Constant elevation check box and specify the elevation instead.
9. Click OK to accept the settings and close the dialog box.
10. Click OK to run the tool.

Compute the camera model

The next step is to run the Compute Camera Model geoprocessing tool to refine the camera model, and adjust images in the mosaic dataset.

1. Open the Compute Camera Model tool.
2. Set the GPS Location Accuracy parameter that best fits your data.
3. Make sure that both Estimate Camera Model and Refine Camera Model are checked.
4. Optionally, you should generate the Output Control Point Table, Output Solution Table, and Output Solution Point Table.

   The Output Control Point Table would be needed to perform additional ground control point processing; this is required if you want to perform the optional refinement steps below. The Output Solution Table and Output Solution Point Table allow you to check the root mean square (RMS) and residuals of the points. For more information about the output tables, see Schemas of block adjustment tables.

Refine the ground control points

Optionally, you can refine the ground control points (GCPs) to produce a more accurate adjustment. To perform this refinement, you will need the Output Control Point Table from the Compute Camera Model tool.

1. Open the Block Adjustment window.
2. Use the Layer list to choose the mosaic dataset that you have just created.

   When you run the block adjustment operations from the Block Adjustment window, you have the advantage of having default naming conventions defined for your outputs. These default names will also be recognized by the Block Adjustment window when specific inputs are needed within the block adjustment workflow.

3. Import the GCPs using the Load GCP button .
4. Click the Block Adjustment Tools list and choose the Compute Block Adjustments tool.
   1. Choose Frame for Transformation Type.
   2. Specify an Output Solution Table.
5. Click the Block Adjustment Tools list and choose the Apply Block Adjustment tool.
   1. Use the solution table output from the previous step as the Input Solution Table.
   2. Click OK to run the tool.

Orthorectify the mosaic dataset with a newly generated DTM

The accuracy of the aerial adjustment can be improved by generating high-resolution elevation data, and then applying it to the mosaic dataset.

1. Use the Build Stereo Model tool to build a stereo table of the mosaic dataset.

   Image pairs with a very small overlap area, a very small intersection angle, or a very large intersection angle should be eliminated while you run the Build Stereo Model tool.

2. Use the Generate Point Cloud tool to generate point cloud data as a folder of LAS files.
   • If your study has less terrain variation, or if you want a quick digital terrain model (DTM) generation, use the ETM (extended terrain matching) method.
   • If your study area has a lot of terrain variation or is an urban area, use the SGM (semiglobal matching) method to generate your DTM or digital surface model (DSM).

   If you want specific pairs to be used in the orthorectification process, open the Stereo table, and set a higher value for those pairs for the Use field. To open the Stereo table, right-click the mosaic layer in the table of contents, point to Open, and click Stereo.

3. Use the Interpolate From Point Cloud tool, using the DTM option to generate a DTM from the folder of LAS files.
4. Use the Apply Block Adjustment tool.
   1. Set the Input solution table using the solution table from the Compute Block Adjustment tool.
   2. Set the Input DEM using the DTM from the Interpolate From Point Cloudtool.