When mapping without Ground Control Points, Photogrammetry data relies on GPS for overall scale and alignment of the Map with the earth. Often this GPS data can be off by up to a few meters and so two maps of the same location, particularly if captured on different days, will not align with each other in 3D space.
When comparing maps for change over time, or re-using blueprint/drawing overlays on multiple maps in the same project, or measuring volume change over time, it's really important that each map aligns with the previous one.
When processing your map, DroneDeploy's Map Engine has a computer vision system that will align any new map you create with previous maps in that location, enabling you to more easily compare them side by side, spot changes, utilise project overlays, and measure stockpile volume change over time.
Features of the new map are compared with features from the previous map. If we find that the maps are similar enough (e.g. most of the scene has not changed), we can usually compute a safe transformation to make the new map align with the previous one.
This means that your maps will align with each other, even without using GCPs.
It also means that if you use Ground Control Points, RTK, or PPK on one map then future maps you create, even without ground control, of the same site will gain much of the benefits of Ground Control automatically, saving you time from laying and picking up ground control on every single flight you do.
Because Map Alignment gives you some of the benefits of GCPs and RTK, it's tempting to use it in situations where provable high-accuracy measurements are required. Unfortunately, the only way to prove your map is accurate is by measuring the location of checkpoints. Either by laying markers on the map in known locations and computing the error, or by uploading with GCPs and checkpoints.
|I want maps that are aligned with design overlays, and to be able to quantify high level changes.||I want to generate high accuracy measurements and make financial decisions directly on the map||I want to have verifiable, proveable accuracy|
|Map Alignment alone will enable you to keep orthomosaics and elevation models well aligned over time. You'll be able to re-use project overlays throughout your project timeline, and compare elevation map to map to identify and quantify change (cut/fill), without needing to use GCP targets or RTK capture on every flight.||High accuracy mapping with DroneDeploy is straightforward. Use Ground Control Points or RTK image capture in addition to Check Points whenever you are making significant financial decisions based on the measurements from the drone data, and be careful to measure appropriately & accurately. Ask your surveyor.||Verifiable accuracy on drone data is a tough standard. Ask your surveyor to review the American Society for Photogrammetry and Remote Sensing (ASPRS) standards from 2015 on this topic.|
We attempt to apply map alignment of every map you process with us, provided that you have previously processed a map in that location. We look to find the most recent map with a similar perimeter and will look back up to five maps to find a map with sufficient overlap. If you would like to see which map your latest map was aligned to, just click the info button next to your map date and read the map date listed beneath "Aligned with previous map". In the example below, the new map on February 25, 2019 was aligned to the previous map from February 4, 2019.
We do not apply map alignment for Live Maps, maps with small percentage overlap with each other, and only when our computed transformation will make the maps more similar to one another.
Yes. While we have very high confidence in the quality of our map alignment technology, we recognise some users may want to opt out of this technology for their own reasons. Please contact our support team to request removal from the Map Alignment workflow.
Updated about a year ago