Appearance
Drone Photos and Photogrammetry
Lidarvisor can build a project from the photos your drone took, with or without a LiDAR point cloud. From the photos it computes a georeferenced, coloured photogrammetric point cloud and an orthophoto, places your ground control points on the photos by itself during the run, and states the accuracy it measured in the project report. The cloud then goes through the same processing as any LiDAR cloud: classification, terrain models, contours, vectors and reports.
Everything happens in one run: you upload, check the photos in the viewer, tick what you need and click Start Processing once.
What You Can Upload
When you create a project, you can give Lidarvisor:
| You upload | What the run can make |
|---|---|
| Drone photos only | A photogrammetric point cloud, then everything you tick on it (classification, DTM, DSM, contours, buildings, trees...), and an orthophoto computed from the photos |
| A LiDAR point cloud and the drone photos of the same site | Your LiDAR results as usual, plus an orthophoto computed from the photos and, if you tick it, a photogrammetric point cloud delivered beside the LiDAR one |
| A LiDAR point cloud only | Unchanged: see Processing Your Data |
Control points, an area of interest and a trajectory can be added to any of these, as described in Creating a Project. If you fly a DJI Zenmuse L1 or L2, you can also upload its raw flight folders and let Lidarvisor compute the LiDAR cloud: see Raw DJI Zenmuse Data.
The Photos
- JPEG photos (
.jpg,.jpeg). Other files in a flight folder (videos,.SRTor.MRKsidecars, and so on) are left out, never refused. - Each photo must carry its GPS position in its file (EXIF), as drone cameras write it. A photo without a position is set aside, with that reason, and the rest of the set goes on.
- Several flights in one project are fine, nadir (camera pointing straight down) and oblique photos alike.
- Photos with an RTK FIX position give the best result. Other photos are accepted: the pre-check then says that the accuracy rests on your control points.
Which Cameras
Any camera whose photos carry a GPS position is accepted. How the camera's lens is calibrated decides how far the accuracy can be trusted, and the photo check and the report say which case applies:
- A camera with a checked profile in Lidarvisor's library (today the wide camera of the DJI Matrice 4T): a calibration checked on known points is used.
- A DJI camera that records its factory calibration in its photos: that calibration is used. Accuracy is not guaranteed.
- Any other camera: it is calibrated from the photos themselves. Accuracy is not guaranteed.
Nothing is refused on these grounds: the run always goes ahead, and the report states what was used.
Creating a Photo Project
- Click Create a project and name the project.
- In Drone photos (folders), pick the folder of each flight. You can add several folders, one after another. Only the JPEG photos are kept. To pick loose photos instead, click Choose JPEG files instead, or Add JPEG files once a folder is in.
- Under the field, Lidarvisor counts the photos, the folders and their size, then reads each photo's position before anything is uploaded ("Reading the photo positions"). It tells you how many photos have no GPS position.
- Optional: pick your LiDAR file in the same panel. A point cloud is optional when you upload photos, but you need one or the other.
- Optional: tick Add ground control points (optional) and pick your control point file. See Control Points on the Photos.
- Click Import project.
Coordinate System of a Photo Only Project
Without a LiDAR file, there is no file header to read a coordinate system from, so the panel shows a Coordinate system block:
- The horizontal system is proposed from the photos (the UTM zone they fall in), for example "EPSG:32633, from the photos. You confirm it when you create the project." You confirm it, or pick another one, in the coordinate system dialog that opens when you click Import project.
- Heights sets what the results' elevations are measured from. It is Ellipsoidal heights by default, which is what RTK drones record. You can pick a geoid height system instead, for example
NAVD88 · GEOID18 · metresnear Houston: only the height systems whose geoid grid covers every photo are listed. See Coordinate Systems and Heights.
With a LiDAR file, the project takes the LiDAR file's coordinate system, as usual.
Upload
While the photos upload, the Photos row of the project tree shows their progress, for example "Uploading 412 of 953 photos". If the upload is interrupted, click Resume sending the photos on that row. When some photos still have no file, the row tells you how many: add the folder or the files they are in, and they are sent.
Photos are added when the project is created. You cannot add photos to an existing project.
After Upload: the Photos in the Viewer
Once the photos are uploaded, Lidarvisor reads them (a few minutes, free of credits): positions, flights, thumbnails, the area they cover, a quality check and, if you uploaded control points, which photos should show each point. Start Processing becomes available when this is done.
The Photos Row
A Photos row appears under Imagery in the project tree. It reads, for example, 953 photos · 2 flights.
- Tick it to show the cameras in 3D: one point per photo, where it was taken, coloured by flight. Pointing at a camera draws the outline of what it looked at. On a photo only project the row is ticked when the project opens.
- Its table button (Photos and cameras) opens the photo panel.
- Its badge gives the result of the photo check.
The Photo Check
Before you process, the photos are checked on their own, and the badge on the Photos row reads Photo check: good, Photo check: worth a look or Photo check: problems found. Its tooltip lists the reasons. It is "Checked before processing, from the photos alone. Nothing here stops the run."
| What the check can say | What it means |
|---|---|
| Fewer than 10 usable photos for a reconstruction. | Too few photos to build a reliable model. |
| Many photos were set aside (no position, duplicate, take-off or landing). | See Photos set aside. |
| No photo looks straight down: the block rests on oblique views only. | Nadir photos are missing. |
| The photos overlap less than usual (front and side percentages given); some areas may not reconstruct. | The flight plan left gaps between photos. |
| Some photos look blurred. | Motion or focus blur on some photos. |
| Not every photo has an RTK FIX position; accuracy rests on the control points. | The drone's positions alone cannot be trusted to the centimetre. |
| No checked calibration for this camera (or the DJI factory calibration is used). Accuracy is not guaranteed. | See Which Cameras. |
| The altitude reference of these photos is not known. | Check the Heights you chose. |
| Some control points are outside the photos (names given). | Those points cannot be placed. |
| The control point heights are some metres away from the photos ground; check their vertical reference. | Usually a height reference mix up (ellipsoidal heights against heights above a geoid). |
| The photos and the point cloud do not seem to cover the same site. | On a project with a LiDAR file: check that you picked the right photos. |
The Photo Panel
Click Photos and cameras on the Photos row. The panel opens along the bottom of the viewer, like the control point and tower panels, and shows the thumbnails grouped by flight, with a count such as "120 of 953 photos".
- Filters: a flight (All flights, Flight 1, ...), All / Nadir / Oblique, Set aside, Left out, and, when you uploaded control points, one point. Points are listed in natural order (1, 2, ... 10, 11). See A Point on the Photos.
- Click a thumbnail to select it and open it large over the 3D view. Beside the thumbnails, the panel keeps a preview of the selected photo with its flight, whether it is nadir or oblique, and its Control points marked during processing; click the preview to open it large again. Fly to the camera moves the 3D view to where the photo was taken.
- In the large view, the arrow keys go to the previous and next photo of the flight, Nearest cameras lists the photos taken nearby, and you can scroll or double click to zoom and drag to move. The full resolution photo loads as you zoom in. Press Esc to close.
Photos Set Aside
Some photos are set aside automatically and never deleted: they stay in the panel under the Set aside filter with their reason:
- no GPS position;
- duplicate of another photo;
- take-off, landing or transit;
- not a JPEG;
- unreadable file.
A photo set aside can be used anyway: select it and click Put back.
Leaving Photos Out
To keep a photo out of the run (a blurred frame, a photo of something else), select it and click Leave out of the run. It stays in the panel under Left out, and Use in the run brings it back. Once processing has started, the panel says "Processing has started: the photos used are fixed for this run." Once the run has finished, the button is gone and the panel says "These are the photos this run used."
Deleting the Photos
Delete the photos on the Photos row removes the photos and their thumbnails from the project and frees their storage. It cannot be undone, and it is refused while the project is processing. Without the photos, nothing can be computed from them any more.
Shared Links
People opening a shared link of your project do not see the photos nor the cameras. The orthophoto and the point clouds stay visible as usual. See Sharing Projects.
Control Points on the Photos
If you upload ground control points with your photos (Add ground control points (optional) in the Create a project panel, the same file as described in Ground Control Points), Lidarvisor places them on the photos for you, during the run. You do not click any mark, and nothing is placed before you start processing.
- Before processing. When the photos are read, Lidarvisor predicts which photos should show each point, from where they were taken. That is only a prediction: no point is marked yet.
- During the run. The photos are first aligned on the drone's positions. Lidarvisor then looks for each target on the photos that should show it. A point is placed when its target is clearly recognised on the photos. The mark is then completed on the other photos that see the point (at least six that agree), and the block is adjusted on the points.
Each point ends in one of these outcomes, all listed in the report: placed, outside the photos (it is in none of the photos, which is not a defect), not found (it is in the photos, but no mark could be confirmed, for example a target that was moved, covered or too small to see), or excluded by you.
Use and Exclude
In the control points panel, each point has a Role. On a photo project, set it before you start processing:
- Fit: the run may use the point. It places it on the photos, then decides by itself whether the point fits the block or checks it (see below).
- Checkpoint: the point is never placed on the photos and never adjusts them. It only measures the delivered cloud, so its figure stays independent.
- Excluded: the run ignores the point entirely. It is not placed and not measured. The report lists it under Excluded by you.
If you set no checkpoint, the run chooses which placed points fit and which check.
Points Used to Fit and Points Kept to Check
To state an honest accuracy, some placed points are kept out of the adjustment and used only to measure the result:
- A Fit point is used to adjust the photos.
- A check point is held out of the adjustment. It measures the accuracy.
- About one point in four is kept as a check point, spread across the site, and never fewer than 3 points are left to fit the block.
- With fewer than 4 placed points, all of them are used to fit, and the report says there is no independent check.
- With at least 4 points used to fit, a point whose position disagrees with all the others before the adjustment (a moved target, a typing mistake) is left out of it, and the report names it with its distance.
The figure that judges the accuracy is the error on the check points. A residual on a point used to fit is never presented as an accuracy.
Points Far from Their Survey
Before the adjustment, every placed point sits at some offset from its survey, and the whole block usually shares most of that offset. A point is far from its survey when its own offset differs from the other points' by much more than they differ among themselves: by more than the largest of 3 times their spread, 10 times the ground sample distance and 10 cm, horizontally or vertically. This usually means a survey or marking error.
- A far point is never used to fit the block. It becomes a check point.
- A far check point is measured and shown with its flag, but it is left out of the headline figure. The headline says so, for example: "1 check point left out of the figure: 0.88 m from where the block and the other points place it, a probable survey or marking error (point 3)."
- If every check point is far, the report says "Accuracy not established from the check points." The verdict is then "Not established", never Error.
- With fewer than 3 placed points, there is nothing to compare with. A point is then kept out of the fit only if it is more than 0.5 m from its survey horizontally.
A Point on the Photos
Pick a point in the photo panel's point filter to see it on the photos.
Before the run, the filter reads Photos predicted to show point and lists the photos that should show it. No mark is drawn, and the preview says: "Control points are marked during processing."
After the run, the filter reads Photos showing point. It lists the photos the run marked, then the photos it examined without finding the mark. Other predicted photos the run never examined are not listed. On each photo:
- a red cross is the run's mark: "Marked during processing";
- a grey ring is the predicted position on a photo the run examined: "Predicted position, no mark found".
The panel also gives the point's status, for example:
- Fit, marked on 14 photos;
- Check, marked on 8 photos, far from its survey;
- Not found: the model was unsure on every photo, followed by the model's own note on what it saw;
- Excluded by you;
- Placed, left out: it disagrees with the other points;
- Outside the photos.
These are the run's own marks, the same ones the report uses. If the control points could not be applied to the block, a placed point reads Marked on 14 photos, not applied: the block was processed without the control points.
For a run that kept no photo list for a point (runs processed before this change), the filter shows the photos predicted to show it, and the panel says: "The run kept no photo list for this point: these are the photos predicted to show it."
In the Control Points Panel
After a photo run, the control points panel shows what the photo adjustment did: a Photo adjustment block above the table, and a Photos column for each point. The block gives the report cover's check point figure and far-point sentence, in the same words, without the 95 % range.
With a LiDAR Cloud as Well
The same control point file also serves the LiDAR cloud: the preliminary check and the Align to control points option work as usual. A point the photo adjustment used is shown in the control points table as "Used to adjust the photos, so not an independent check", and is kept out of the accuracy figures.
On a photo only project, there is no cloud before processing, so the control points panel waits: "Your control points will be checked against the point cloud once it has finished processing."
Processing Options with Photos
With photos in the project, the processing panel opens with a Photogrammetric Point Cloud Options category, and the Imagery Options category computes the orthophoto from the photos:
| Category | Option | What it does |
|---|---|---|
| Photogrammetric Point Cloud Options | Compute photogrammetric point cloud | Builds a coloured point cloud from your photos. On a photo only project it is the cloud every other option runs on, so it is ticked by default. With a LiDAR cloud as well, it is unticked by default; when ticked, the photogrammetric cloud is delivered beside the LiDAR one, and it is not classified. |
| Photogrammetric Point Cloud Options | Photogrammetric cloud quality | Shown under the option above while it is ticked: Low, Medium (default) or High. See below. |
| Photogrammetric Point Cloud Options | Compute 3D mesh | A 3D model of the site textured with your photos, with buildings and trees. View it in 3D in the viewer; download it as OBJ with textures (.zip) or GLB. It is built from the photogrammetric point cloud, which is computed with it. Off by default. Mesh detail: Low, Medium (default) or High. |
| Photogrammetric Point Cloud Options | Compute Gaussian splat | A photo-real 3D view of your site, best for buildings, glass, trees and fences. It is for viewing: the measuring tools do not pick on it (measure on the point cloud or the mesh). Download as PLY (.zip) for SuperSplat, Postshot, Cesium ion, Unreal or Blender plug-ins. Looks best when your flight includes oblique photos. Off by default. Splat detail: Low, Medium (default) or High. |
| Imagery Options | Compute orthophoto from your photos | Assembles your photos into one orthophoto. Roads, road markings, water and rails read it. It replaces the Generate Orthophoto option when the project has photos. |
On a project with LiDAR and photos, the next category is titled LiDAR Point Cloud Options: its options (classification, alignment, thinning, smoothing...) apply to the LiDAR cloud. On a photo only project it keeps its usual name, Point Cloud Options, and its options apply to the photogrammetric cloud.
If you uploaded your own orthophoto, it is kept and no orthophoto is computed: Compute orthophoto from your photos is then not offered.
Without the photogrammetric point cloud, the orthophoto is computed on a simplified surface: right on open ground, but buildings and trees may lean. Tick both for the best orthophoto.
The mesh and the splat add time to the run, never credits: the price is the same with or without them, and at every detail level.
On a project with a LiDAR cloud as well (or a raw DJI flight), the orthophoto is projected on the LiDAR surface rather than on a surface computed from the photos, so buildings and trees stand where the LiDAR measured them instead of leaning. The run measures the orthophoto against the LiDAR and warns when the two disagree.
Colorize Point Cloud on a LiDAR and photos project colours the LiDAR cloud from your photos, and from satellite imagery where no photo covers it. It applies only when the LiDAR cloud carries no colour of its own.
Photogrammetric Cloud Quality
The quality sets how finely depth is computed in each photo. The tooltip gives, from your own photos, the size on the ground of one depth map pixel at each level: each level up halves it, for more detail on edges and small objects.
| Level | Effect |
|---|---|
| Low | Coarser cloud. About a quarter of the time of the dense cloud step. |
| Medium (default) | The balanced setting. |
| High | Finer cloud. About four times the time of the dense cloud step. |
The price is the same at every level. If the High level runs out of memory on a very large set of photos, that step is computed again at Medium, and the report says so.
What Changes on a Photo Only Project
- Compute photogrammetric point cloud is the base of everything. If you untick it, every option that needs a point cloud is greyed out, with a line saying that these options need a point cloud, and only the orthophoto and the project report remain. With neither the cloud nor the orthophoto ticked, there is nothing to process.
- The Photogrammetric Point Cloud switch under classification is on and locked: the cloud comes from photos by construction.
- The power line options and the vegetation encroachment report are not available: photogrammetry cannot reconstruct the conductor wires.
- Align Flight Lines is not offered: a cloud built from photos has no LiDAR flight lines, and the photo adjustment already makes the photos agree.
- Align to control points is not offered: your control points are already used in the photo adjustment, and the cloud is checked against them in the report.
- Thin Point Cloud applies to the photogrammetric cloud (its tooltip says "Applies to the photogrammetric point cloud."). On a project with LiDAR, it applies to the LiDAR cloud.
During the Run
When the project has photos, the photo steps come first, so that the orthophoto from your photos is ready for the detections that read it. The progress line names each step:
- Aligning the photos...
- Placing the control points on the photos... (with control points only)
- Computing the point cloud and orthophoto from the photos...
- Georeferencing the photogrammetry results...
Then the rest of the run continues as for any cloud. You can close your browser: you receive an email when processing finishes.
Results
Depending on what you uploaded and ticked, you get:
- The photogrammetric point cloud. On a photo only project it is the project's point cloud, classified and used for every product you ticked. On a LiDAR and photos project it appears under Imagery as Photogrammetric point cloud, beside the LiDAR cloud.
- The orthophoto from your photos, viewable as an image layer and downloadable as a GeoTIFF.
- The 3D mesh and the Gaussian splat, when you ticked them: each is a layer of the tree, shown in 3D, and downloadable (OBJ or GLB for the mesh, PLY for the splat). OBJ and GLB coordinates are relative to an origin stated in the download, with a
.prjnaming your coordinate system. - The terrain models, contours and vectors you ticked, built by the same rules as for a LiDAR cloud. Remember that a photogrammetric cloud only captures what the camera saw: there is little or no ground under dense vegetation, so the terrain there is interpolated.
- The accuracy section of the project report (below).
The photogrammetric point cloud, the orthophoto computed from your photos, the mesh and the splat are processed results: downloading them needs a paid plan, like the classified point cloud. The photos themselves stay an input of the project and are not downloaded from the tree. See Downloading and Exporting.
The Accuracy Report
When photos were used, the project report (the Generate Project Report option, on by default) gains a Photogrammetry and absolute accuracy section, and its cover describes the job as a drone photogrammetry survey when there is no LiDAR cloud.
On the Cover
On a project with photos only, the cover gives two accuracy figures, each with its name, and a verdict. They measure two different things:
- Check points (marked on the photos): the horizontal and vertical error (RMSE) measured on the check points, the number of check points, and the 95 % range in which the true figure lies given that number. It measures the photo block in 3D. With 1 check point the range is very wide (for example 11 to 134 cm for a measured 21 cm); with fewer than 3 the report says plainly that there are too few check points to establish the accuracy. A far check point is left out of this figure, and one sentence names it. If every check point is far, the line reads "Accuracy not established from the check points."
- Cloud heights at the control points: for example "RMSEz 7.7 cm on 9 points (surveyed minus cloud)". It compares the delivered cloud with your surveyed heights, vertically, at every point not used to adjust the photos. It is the same check as the control points panel. It appears only when that check was measured.
- Accuracy verdict: pass, warning or error, measured on the check points in multiples of the ground sample distance (up to 1.5 times passes, up to 2.5 times is a warning), with one sentence saying what it means. When every check point is far, the verdict is "Not established: every check point is far from its survey".
The two figures often differ: the first is a 3D test on a few check points, the second a vertical test on more points. Neither replaces the other.
In the Section
The headline, the same as the cover's first line, and one sentence saying how it differs from the section Cloud heights at the control points.
The accuracy statement, worded as the ASPRS Positional Accuracy Standards (2nd edition) require. The standard asks for at least 30 check points; with fewer, the report says it was tested "using only N checkpoints".
Warnings, each in one sentence:
- the control points are fitted much better than the check points (one may be wrong);
- a check point stands out from the others as a probable blunder;
- the check points share a statistically significant offset (a bias).
Check points and fit points in two tables, one row per point. Each row gives:
- the error east, north and up (the point measured in the delivered block minus its survey);
- the horizontal error;
- the reprojection error in pixels;
- the number of photos;
- for a check point, its distance to the nearest fit point.
Each table ends with the mean (bias), the largest error and the RMSE. A fit point's residual is never presented as an accuracy: it was used in the adjustment.
A far check point stays in the check table, marked with an asterisk: "Far from where the block and the other points place it before adjustment: a probable survey or marking error. It is measured and shown here, and left out of the headline figure." The table then also gives the RMSE without far points, and the section adds a Check RMS with every check point (far points included).
Control points not used, with the reason for each (for example: no mark found on the photos, found on one photo only, the marks found did not meet in one point, outside the photo block), followed by the model's own note when it gave one.
Excluded by you: the points you excluded in the control points panel, for example "Points you excluded in the control points panel: 4, 9. The run neither placed nor measured them."
A map of the control points on the orthophoto. It distinguishes check points and fit points, draws each horizontal error as an arrow with its scale stated, and shows the vertical error as a colour.
Survey and processing:
- the share of photos with an RTK FIX position and their accuracy;
- how the photos were aligned;
- the overlap;
- the ground sample distance;
- the camera calibration: the focal length after the adjustment and its change from the reference. When the camera calibrated itself from the photos and the photos carry the maker's calibration (as DJI cameras do), that is the reference, for example "+0.77 % from the maker's calibration (3,666.7 px); started from 3,242.7 px". When a calibration was imposed, the change is given from the imposed calibration;
- the offset of the drone's positions from the control points before adjustment;
- how the control points entered the adjustment;
- the height reference (for example ellipsoidal heights, not altitudes above sea level).
The photo block: how many photos were supplied and aligned, the nadir cameras expected and found, the reprojection error and its verdict.
The pre-check of the photos, with its reasons.
The photos each control point was marked on are in an annex at the end of the report: one sheet per point, with the mark drawn on every photo, so you can see what was measured. A point that was not found has a sheet too: the most vertical photos the run looked at, cropped where the point should be, with a grey ring there. You see at a glance whether the target was on the photos.
Cloud Heights at the Control Points
On a project with photos only, the report's control point section (the Absolute accuracy section of a LiDAR project, see Ground Control Points) is titled Cloud heights at the control points. It compares the delivered cloud's heights with your surveyed heights (surveyed minus cloud), with the ASPRS vertical statement. The points used to adjust the photos are named and left out of it: "these figures use only the points the adjustment did not". Its Correction row says the cloud comes from photos adjusted on your control points, and names them.
The report always says what it could not establish:
- With no control points: "No control points: absolute accuracy = the drone's RTK only."
- With no check point measured: there is no independent check, and it says so.
- If the control points could not be applied to the block, it was processed without them, and the report says that its accuracy is that of the drone's positions only.
No run is refused or failed on accuracy. A block that went wrong is delivered with a red verdict that says in plain words what happened.
Credits
Photogrammetry follows the usual rule: 1 credit per hectare, with a minimum of 5 credits per run, one charge per run, whatever you tick. The area depends on what the project holds:
| Project | Area charged |
|---|---|
| Photos only | The area the photos cover (their footprint, measured when the photos are read). An area of interest crops the results but does not lower the price, because every photo is processed. |
| LiDAR and photos, with a photo product ticked | The larger of the LiDAR area (or of your area of interest inside it) and the area the photos cover. |
| LiDAR and photos, no photo product ticked | The LiDAR area, as for any LiDAR project. |
The area is counted in whole hectares, with a minimum of 5 credits per run. The confirmation dialog shows the exact number before anything is charged. If the run fails, the credits are refunded automatically, as for any run. See Account and Subscription.
Limits and What Is Not Supported
Photos per Plan
Each plan computes up to a number of gigapixels of photos per run: the pixels of every photo the run uses (width × height), added up. Photos you set aside or left out do not count.
| Plan | Photos per run |
|---|---|
| Freemium | 4 Gpx, about 200 photos at 20 MP |
| Lite | 20 Gpx, about 1,000 photos at 20 MP |
| Premium | 60 Gpx, about 3,000 photos at 20 MP |
| Advanced | 100 Gpx, about 5,000 photos at 20 MP |
The processing panel shows where your project stands, for example "1,245 photos · 24.9 Gpx of 20 Gpx". Above the limit, Compute photogrammetric point cloud and Compute orthophoto from your photos are greyed out, with "Your plan computes up to N Gpx of photos (about M photos at 20 MP). Exclude photos or upgrade.", and the rest of the run stays available. To go ahead, leave photos out until the set fits, or upgrade. A set exactly at the limit is accepted. The conversion of raw DJI data is never limited this way.
What Is Not Supported
- Photos are added at project creation. They cannot be added to a project later.
- A project is processed once, as any project: tick every product you need before you start.
- JPEG photos with a GPS position only. Thermal and multispectral bands, video frames, and handheld or terrestrial photos are not supported.
- The LiDAR and photogrammetric clouds are not merged: on a LiDAR and photos project, the LiDAR cloud is the one that is classified, and the photogrammetric cloud is delivered beside it.
- What a camera cannot see is not reconstructed: the ground under dense canopy, thin wires, the inside of shadows.
Next Step
Drone LiDAR flown with a DJI Zenmuse L1 or L2? Upload its raw folders directly: head to Raw DJI Zenmuse Data. Otherwise, see Processing Your Data for every other option.