March 12, 2024

Automatic Identification of Ground Control Points in Drone/UAV Imagery

Written by: Amanda Lind

 

Ground Control Points (GCP) are an important method for improving accuracy in photogrammetry. This process has greatly improved in Global Mapper Pro v25.1. The benefit of including GCPs in processing is that it allows users to leverage the precision of high-accuracy GNSS units to imagery collected via drone/UAV. Traditionally this has been a painstakingly manual workflow, as each point feature must be associated with the pixel in which it appears throughout the image set. To streamline this, a new tool in Global Mapper Pro v25.1 automatically identifies and marks GCPs in images making this process faster and less time-consuming.

A screenshot of the ground control point manager
The ground control points manager is the new tool for automatically placing GCP markers in your imagery.

What is Pixels to Points (P2P)?

Pixels to Points, often referred to as P2P, is a photogrammetric processing tool within Global Mapper Pro that uses overlapping images to create photo-textured 3D models, orthoimages, and high-resolution point clouds. These generated data models can then be used in Global Mapper to measure terrain volume, assess structures, create detailed base maps, and more.

The Ground Control Points manager is the latest workflow improvement tool in Pixels to Points (P2P). This is a part of Blue Marble’s ongoing work to improve P2P output quality and processing speed with each version release. Learn more about other recent updates in P2P in this blog: Improvements to the Pixels to Points tool in v25.0

The New Ground Control Points Manager 

New advanced Ground control point functionalities in Global Mapper Pro v25.1 have been organized into a dedicated management window. The GCP manager can be accessed from the main P2P dialog. First, load the images, then choose the Ground Control Points button to open the manager. Here, click on a point from the list to populate the suggested images list with images that overlap with the selected point. This list is based on an approximate measure of coverage, as the images’ footprints aren’t officially calculated until the photogrammetric process begins. The automatic GCP identification tool can recognize many different types of unique GCPs in the imagery, as further discussed below. For GCPs that cannot be visually identified on their own, manual point placement can be done for each image using the Add Control Point to Image option.

Automatic GCP Identification and Placement 

Functionally, the basic requirement for a ground control point target is that it must be a unique object that is clearly identifiable in the imagery. While a traditional survey marker is ideal, the corner of a parking line, a paper plate with a duct-taped X on it staked to the ground, something that you can see and mark the precise center of, should work as well. If the features are unique, they can be automatically placed with the new GCP tool. This process uses a method of shape detection to identify similar features in the other images.

Four different ground control points, as described in the caption.
Some examples of unique, identifiable GCPs include spray-painted X’s, intersecting lines of a tennis court, a spray-painted marker in the grass and more.

Use the Define Target in Image button to identify the ground control point in an image. Click on the center of the target (where you want the control point to be placed) and drag your mouse outwards until the box covers the target. The feature will be placed at the center of the target, as marked in the image. Zoom in and try to be as accurate as possible when choosing the pixel to represent the center of the GCP. This will be used to identify this GCP in the overlapping images and only needs to be done once per GCP.

A screenshot of the populated ground control points manager
Images will be color-coded based on the likelihood of point identification.

Based on its assessment, Global Mapper will color code the images based on its confidence in point identification. Images with Confirmed Control Point locations that you have approved will be highlighted in blue. Auto-placed markers that have not been verified are green, and images that have Possible Overlap with the GCP but don’t contain a target that could be visually identified are yellow. To improve accuracy to the best extent, confirm as many GCP locations as possible. Only confirmed point locations will be included during processing. Unconfirmed auto-placed will be ignored. 

To confirm a GCP location, click on an image from the list and locate the target in the image. If the point is correctly placed in the image, confirm its location by checking the box next to the image name.  If necessary, adjust the control point location by clicking and dragging the point feature. Adjusting the location will automatically confirm the image. Note: point symbol colors and labels can be changed by double-clicking on the ground control point from the list. For full step-by-step instructions, check out the Knowledge Base.

Placed markers can be saved for future processing. This option can be found in the main P2P window by exporting a .P2Pgmp from the File menu. This will include your images, GCP layer, processing and wizard settings, and GCP positions in the images.

The GCP manager and automatic GCP identification are just a few updates available in Global Mapper Pro v25.1. Read about additional new features in Top New Features in Global Mapper Pro v25.1.

To check out the exciting new functionality in Global Mapper standard and Pro v25.1, download a free 14-day trial today!  If you have any questions, please contact us!

Companies using Blue Marble’s geospatial technology