October 31, 2023

Improvements to the Pixels to Points tool in v25.0

Written by: Amanda Lind

 

Pixels to Points is a photogrammetric processing tool within Global Mapper Pro that creates photo-textured 3D models, orthoimages, and high-resolution point clouds via Structure from Motion (SfM) and Multi-View Stereovision. Using overlapping UAV-collected images, these generated data models can be used in Global Mapper to measure terrain volume, asses structures, create detailed base maps, and more. 

The new version 25 release provides yet more improvements to the Pixels to Points tool, including significant speed advancements to decrease the processing time and revisions to the Wizard to streamline your workflow further than before. These improvements came from extensive fine-tuning in development and QA, and through Beta testers for the software. Typically a few weeks before each release, a new version of the software is provided for Beta testers who volunteer to try out the new software. In exchange for early access to new tools and updates, these Beta testers help contribute to the process that makes Global Mapper such an outstanding software.

A photogrammetrically generated model of a railroad through a neighborhood
Pixels to Points is an easy tool for generating accurate, high-resolution 3D data for your area of interest.

How does Pixels to Points work ‘Under the Hood’?

A model of how overlapping images are used in Structure from Motion
The overlap between images is what provides the information necessary to create 3D data from 2D imagery.

Let us take a broad look at the processes that make up Pixels to Points. Structure from Motion is used to process the images and find features:

  • First, the camera information is computed. Typically, each image has x,y, and z information from the drone stored as EXIF data, but the software needs to calculate additional parameters to determine how the image plane is related to the ground/world. This is part of why images initially load as picture points in the software, Global Mapper knows the camera was when the image was taken but the image is not yet fully georeferenced
  • Next, it finds features in the images. These features, such as the corner of a building or a large rock in a field, etc. are represented by distinguishable patterns of pixels. By identifying the same feature in multiple images, its 3D location can be triangulated. This process is similar to how human brains calculate depth perception using the change in perspective between our two eyes.
  • The calculated position of these features is then related to the real-world coordinates provided by ground control points or EXIF data from the images. 

For more information check out this blog: How Pixels to Points Works

The GM P2P process uses a series of steps across two separate libraries: OpenMVG and OpenMVS. Speed improvement for this recent update focused on the most time-consuming steps of this process including SfM in OpenMVG. These speedups in v25.0 are all CPU-based and do not require a GPU, although the whole process runs better with a GPU. 

Updates to the Wizard 

The Pixels to Points Wizard is designed to increase ease of use for existing users and make photogrammetric processing more approachable to new users by streamlining settings choices. Simply import your images and choose your most important output, including the option to run a quick output for quality assessment. Global Mapper adjusts the Pixel to Points settings to match your input, providing an optional follow-up option to tweak the settings if desired. Use the primary dialog to add control points and assess more advanced settings such as masking and color harmonization. With Global Mapper Pro’s powerful point cloud processing tools, Pixels to Points provides a powerful and streamlined option for creating and editing your spatial data.

The Pixels to Points Wizard
The Wizard automatically tailors processing settings based on your goal output.

Just like all tools in Global Mapper, there are many smaller updates to help make sure users get the best outputs. One of these updates to the wizard that many users may not often encounter is the Projection Advisor. This guided warning will trigger when a user applies a projection to the output that may not be compatible with the source images. For example, if the images were collected in The US, but the user choose a Polar projection. This is another step of preprocessing output validation that happens behind the scenes in Pixels to points to save you time and effort. 

If you’re interested in becoming a Beta tester, email Beta@bluemarblegeo.com or keep an eye on the Global Mapper Forum page. 

To try these new updates with your data, download a 14-day free trial of Global Mapper Pro v25 today! If you have any questions, please contact us!

Companies using Blue Marble’s geospatial technology