When an ortho mapping workspace is created, the next step is to perform block adjustment using the tools in the Adjust and Refine groups. The adjustment tools encapsulate operations of computing match points (tie points) from overlapped images and performing triangulation calculations. The adjustment tools use an algorithm and processing procedure that are suitable for each ortho mapping workspace type. For instance, the algorithm performs RPC triangulation for satellite images and frame triangulation for aerial images. The parameters used in the adjustment computation can be defined on the Adjustment Options dialog box. The adjustment quality report can be viewed using the Quality Report tool.
The Adjust tool is used to perform block adjustment of the image collection in the current ortho mapping workspace. The processing is always done at image source resolution—computing tie points from source images, and then performing triangulation using the computed tie points. Drone and scanned aerial images normally come with poor-quality or no exterior orientations and coarse GPS information. In these cases, an initial adjustment will be performed at a coarse resolution—a user-defined pyramids level is used to estimate the image orientations, and then an adjustment will be performed at the source resolution. The tie points generated at the initial adjustment will be used to seed the tie point computation and triangulation calculation at the source resolution. Having pyramids built for the image collection ahead of time or during ortho mapping workspace creation will reduce the time of initial adjustment process.
The Quick Adjust tool is applicable for processing drone images. It performs adjustment at a coarse resolution to estimate exterior image orientations and improve GPS location. This initial adjustment is done quickly and allows you to get an understanding of the data coverage of your study area and the choice of the processing parameters for this collection before taking the time to run an adjustment at source resolution. For example, when you collect data in the field, you can run this tool and run Adjust again to compute the full adjustment. This can also be used as a step to establish the camera model before adding GCPs.
To further improve the adjustment accuracy with ground reference, you can add ground control points (GCP) using the Manage GCP tools. The adjustment quality can be viewed from an adjustment report using the Adjustment Report tool. If the adjustment errors are larger than expected, there are several ways you can check your tie points:
- Check the solution point layer in the Content pane, review the residual map or analyze the residuals in the Residual field in the attribute table, identify point IDs that have large residuals, and delete tie points that have the same point IDs in the control point table.
- Run the Analyze Tie Point tool to generate a coverage feature class and an overlap polygon feature class. The two feature classes help you understand whether you have enough tie points and where you need to add more tie points.
- Use the Recompute Tie Points tool to compute tie points and to also append more tie points.
These tools will update the control points in the ortho mapping workspace. You can run the Adjust tool again to perform triangulation and update the image collection.
Add GCPs to your ortho mapping workspace
You can enter GCPs into your ortho mapping workspace by either importing them or manually adding GCPs based on a reference layer in the map.
The Import GCPs tool can be used to import GCPs stored in a .csv file or a text file. The values should include GCP name, X coordinate, Y coordinate, Z coordinate, Horizontal accuracy, and Vertical accuracy. The spatial reference of the coordinates should be provided while importing the GCPs. Sometimes GCP data comes with photos of the location where the GCPs were measured. If there are multiple photos for each GCP, the GCP label should include number which will be used as a unique identifier. For example GCP20_ov.jpg, GCP20_1.jpg, GCP20_2.jpg are the photos for GCP 20.
After the GCPs are imported, click the Manage GCPs tool to open the GCP Manager pane, which allows you to add the corresponding tie points for these GCPs. Click the GCP point in the list, and corresponding images that overlap with the GCP show up in the image list. Click the Add GCP or Tie Point button to add an image tie point in the viewer. The image points for other images will automatically be added by the image matching algorithm when possible. The Clear Links button is used to delete tie points from images, and the Delete GCP button is used to delete GCPs and the associated tie points from image viewers.
If you don't have a GCP file and you need to measure GCPs on a reference map, use the Manage GCPs tool to enter a GCP by clicking a location on the map, and then add the corresponding tie points in the viewers.
If you have a georeferenced raster layer (raster dataset, mosaic dataset, or image service) that has a resolution similar to your image collection, you can add it as a reference to compute GCPs using the Compute Control Points tool.
To learn about creating ortho mapping products, see Generate ortho mapping products.