Using drone data in Sphere XG is currently in beta state. As a beta release, this feature is designed for early access for all Sphere XG customers with a Point Cloud subscription and will continue to evolve, shaped by real customer feedback and ongoing improvements.
The quality of your drone data capture is largely determined before the drone leaves the ground. Usually, poor quality in the created point cloud is not due to the software, but to the way the data was captured.
Typical issues are:
Insufficient photo overlap
Inconsistent flight height
Blurred images due to motion
Poor lighting
Incorrect camera orientation
In other words, your project start with good flight planning. You can find an overview of best practices in the following sections:
Walk the Site
Before you start planning the drone flight and preparing the hardware, you should walk the site you want to capture. Take note of reflective surfaces, repetitive textures, water, cranes, traffic, moving vegetation, and areas obstructed by GNSS, i.e. areas where GPS reception is poor, for example near tall buildings, cliffs or dense trees.
Run Pre-Flight Checks
Clean the camera lens to avoid blurry images.
Check the data storage to make sure it is enough for the full mission.
Verify the correct coordinate reference system (CRS) and geoid settings.
Check the photo capture interval for a good overlap.
Enable geotagging to record accurate position data for each photo.
Apart from these checks, you should take note of the project requirements regarding data budget, expected number of photos, estimated total data volume and intended export format. This will help you to plan the project in one go or split into phases.
Select the Right Flight Pattern for Your Site
Any drone flight should start and end outside the area you want to capture. You can find illustrations of standard flight patterns below.
Open terrain (planar mapping): For example, open ground, roads, stockpiles, earthworks, etc. For this type of terrain, use a regular grid. This makes sure that there will be consistent forward and side overlap in the drone photos.
Terrain with height variations: Try to keep the altitude above ground as constant as possible. In other words, if the terrain rises, adjust the altitude above ground. This makes sure that the image scaling and the GSD* remain as consistent as possible.
Terrain with complex relief: For example, loose material, irregular stockpiles, etc. Use a double grid or a crosshatch flight pattern.
Site with vertical surfaces: For example, buildings, embankments, walls, towers, quarries. In this case, you need photos taken from the side and from above. Plan flight paths where the camera looks sideways and at an angle, not straight down.
| 
| 
|
Single grid | Double grid survey | Circular flight mission |
If you want to check the capture quality, use checkpoints or a known distance in the scene.
* Ground Sampling Distance (GSD) describes the size of the area on the ground represented by a single pixel in a photo and indicates the level of spatial resolution and visible detail in photogrammetric outputs.
Select Altitude and Speed that Fit Your Project
Make sure that the drone flight does not lead to dangerous situations where people or property could be harmed.
Respect local regulations.
Check the result requirements to decide the altitude. Flying at a lower altitude provides more detail, a denser point cloud but also results in a higher data volume. Flying at a higher altitude is more efficient regarding fast coverage of a site and reducing the data volume but provides a less dense and less detailed point cloud.
Select the right speed based on the equipment: If the camera has a fast shutter (mechanical or global), you can fly at a higher speed. If the camera has rolling shutter sensors, you should fly at a lower speed.
If you have to fly in conditions that include wind, vibration or bad lighting, select a lower speed.
Camera and Lighting Basics
Use a camera with reliable GPS geotagging.
Your drone must have GPS and record a GPS position for every photo. Photogrammetry can work without geotags, but alignment is slower and more error‑prone. In addition, it might be more difficult to trace images back to their location in the field. Note also that you cannot upload photos without GPS position to Sphere XG.Choose the right shutter.
It is recommended to use mechanical or global shutters for mapping because they capture the whole image at once. Rolling shutters can still work, but they are more sensitive to flight speed, vibration, and motion.Keep a stable focus throughout the flight.
You should not try to reach the best autofocus constantly during the mission. Consistent sharp images are more important than fine‑tuning the focus for individual scenes.Keep balanced exposure settings.
Avoid extreme underexposure, overexposure or deep shadows. Photos that are too dark, too bright, or full of deep shadows do not capture enough detail. Photogrammetry relies on visible surface texture.Use consistent white balance and image settings.
Sudden changes in color or contrast between photos make subsequent processing harder. You should avoid aggressive automatic image enhancements that vary from photo to photo.Fly in even lighting conditions when possible.
Bright overcast conditions are ideal. They reduce harsh shadows and reflections, leading to cleaner and more consistent results.Take note of weather conditions.
Note any weather conditions that may affect later processing behavior, such as wind, sun angle, any site particularities.Check the capture before leaving the site.
Review a sample of photos in the field before leaving. That way, you can identify blur or exposure problems immediately and correct them before you are back in the office.
TIP: If you want an independent reality check, you can place a scale bar or any other object with a known length in the scene. The scale bar or object does not need to be in every photo, but it must be visible in overlapping photos with good geometry. This allows you to verify whether the reconstructed scale matches a known length.
GPS, Accuracy, and Confidence
GPS on the drone provides an approximate position for each photo. This is useful for organizing the site, speeding up alignment, and establishing an initial scale. However, these photo geotags do not guarantee survey‑grade accuracy. Final results depend on factors like sensor quality, GNSS performance, flight geometry, scene texture, and whether you use independent checks.
The table below shows an overview of capturing methods including their strengths, limitations and accuracy.
Method | Strength | Limitations | Accuracy |
RTK* or PPK** photo positions | Improves camera position quality and often reduces the amount of ground control needed. | Still requires proper mission design and independent validation where accuracy matters. | 1-10 cm |
Photo geotags / onboard GPS | Fast, automatic, good for initial block placement and first-release scaling. Cheaper than RTK drones. | Accuracy is limited by onboard positioning quality and should be described accurately. | 1-10 m |
Ground control and checkpoints | Best for stronger georeferencing and transparent accuracy reporting. | Needs field survey effort and a disciplined control/checkpoint split. | Depends on selection accuracy |
Known scale bar / known distance | Quick sanity check on model scale and deformation. | Not a substitute for full georeferencing across a large site. | Depends on user selection accuracy |
*RTK (Real-Time Kinematic): a GNSS positioning method that improves a drone’s GPS accuracy from a few meters to centimeter-level precision by using corrections from a nearby reference station in real time.
**PPK (Post‑Processed Kinematic): a GNSS method that improves position accuracy after the flight by applying correction data from a reference station during data processing.
Why Do We release a Beta Version?
We are excited you are part of this journey!
Our goal is to provide our customers with reliable, high-quality solutions for their reality captures. We also want to develop our solutions with a strong focus on your requirements. That's why we've decided to roll out our first photogrammetry solution in a beta state with a scope that is currently limited. Based on your feedback and findings, we will strive to expand and streamline the Sphere XG photogrammetry.
But we also want to be very clear about what this beta version can and cannot do. So, keep the following limitations in mind when you plan and capture your project.
No user‑defined scaling during processing
You cannot adjust the scale after uploading. While the visual representation may be accurate, measurements rely on capture metadata (GPS, RTK, or PPK) and external validation only.GPS‑based scale only
Results are scaled from photo positions, not survey control. This works for visualization and rough measurements, but it is not survey‑grade accuracy.Single‑flight capture recommended
Capture the site in one flight with one drone. Otherwise, height inconsistencies of 10 to 20 m might be the result.No access to images after upload
It is currently not possible to view or retrieve uploaded images. Review, name, and archive imagery before upload. Make sure to keep master copies of the source imagery, flight logs, and field notes outside the processing platform.Limited export formats
Exports are currently limited to .las, .laz and .e57 point clouds. Ortho mosaic and other imagery outputs will be available in the near future.
Quick Checklist for the Field
Flight pattern selected to match the site, i.e. grid, double-grid, corridor, circular, or mixed.
Forward and side overlap are intentionally set higher for difficult surfaces and relief.
Drone GPS is active and image geotagging has been confirmed.
Camera and lens cleaned. Focus and exposure checked.
Altitude connected to target GSD and legal limits.
Speed defined depending on image sharpness and sensor type.
Scale bar, checkpoints, or other validation method prepared if needed.
Data budget reviewed for storage, upload, and platform limits.
Sample images reviewed prior to leaving the site.