LiDAR on drones: 3D mapping within everyone's reach

In this article

TL;DR:

LiDAR on drones enables 3D models with centimeter-level accuracy: for the DJI Zenmuse L2, DJI specifies 4 cm vertical and 5 cm horizontal at 150 m altitude
The DJI Zenmuse L2 democratized access to aerial LiDAR, priced from €10,500
LiDAR can measure the ground under dense vegetation, through the gaps in the canopy, something impossible with conventional photogrammetry
For open terrain and buildings, photogrammetry remains more economical and sufficiently accurate
The LiDAR + photogrammetry combination (hybrid sensors) is the dominant trend in 2026

For years, aerial LiDAR seemed reserved for big engineering projects and well-funded universities: very expensive sensors, post-processing that demanded dedicated computers, and a learning curve that discouraged any mere mortal.

That has changed. In 2026, an operator who already has a compatible drone can do LiDAR 3D mapping with a sensor in the €10,000 to €15,000 range. The question is no longer “can I use LiDAR?” but rather “should I use LiDAR or photogrammetry for my project?”

How LiDAR works (without overcomplicating it)

LiDAR stands for Light Detection and Ranging. The principle is simple: the sensor emits thousands of laser pulses per second and measures the time each pulse takes to return after hitting a surface. With that time, it calculates distance with millimetric precision. Multiply by hundreds of thousands of points per second, and you get a three-dimensional “point cloud” of the terrain or object.

For those accustomed to photogrammetry (which uses overlapping photographs to reconstruct 3D geometry), the key difference is that LiDAR directly measures geometry — it doesn’t estimate, it measures. This translates to higher accuracy and, crucially, the ability to “see through” vegetation.

The multiple returns phenomenon

This is the feature that makes LiDAR truly special for certain applications. When a laser pulse hits a tree canopy, part of the energy is reflected by foliage (first return), but some continues penetrating until it hits branches, trunks, and eventually the ground. Modern sensors record up to five returns per pulse.

In practice, this means you can map terrain under dense forest without cutting a single tree. For those working in forestry, archaeology, or infrastructure planning in forested areas, there’s no viable alternative — photogrammetry simply cannot see the ground under closed canopies.

DJI Zenmuse L2: the game-changing sensor

I’ll be direct: the DJI Zenmuse L2 is currently the best value for money in drone LiDAR. I’m not saying this because I’m a DJI fan (I have my criticisms), but because the numbers speak for themselves.

Specifications that matter:

Point rate: 240,000 points/second in single-return mode; up to 1.2 million/second with multiple returns (up to 5)
Range: up to 450 meters at 50% reflectivity (250 meters at 10%)
Vertical accuracy: 4 cm at 150 m altitude (5 cm horizontal)
Integrated RGB camera: 20 MP with stabilization
High-accuracy IMU: together with the drone’s RTK, georeferences the points with centimeter-level accuracy
Weight: 905 grams, compatible with Matrice 400, Matrice 350 RTK and Matrice 300 RTK (the latter with the DJI RC Plus controller)
Price: from €10,500 (sensor) + Matrice 350 RTK drone (~€8,900) = ~€19,400 complete

The L2 hit the sweet spot between capability and price.

Typical L2 workflow

A typical LiDAR survey with the Zenmuse L2 follows these steps:

1. Flight planning: Using DJI Pilot 2 or DJI Terra. Typical altitude: 80-100 meters. Speed: 8-12 m/s. Side overlap: 30%.

2. Flight and capture: At 240,000 points/second, a 30-minute mission can generate hundreds of millions of points and several gigabytes of raw data.

3. Post-processing: DJI Terra generates the classified point cloud (ground, vegetation, buildings). For complex projects, there’s CloudCompare (free) or Terrasolid.

4. Deliverables: DTM, DSM, contour lines, profiles, and RGB orthomosaic. All in standard formats (LAS/LAZ, GeoTIFF, DXF).

LiDAR vs. Photogrammetry: when to use each

This is the golden question, and the answer is more nuanced than most equipment salespeople want to admit.

Use LiDAR when:

Dense vegetation: If you need the terrain model under forest or scrub, there’s no alternative.
Critical altimetric accuracy: For civil engineering projects with sub-5 cm tolerances.
Large areas with tight deadlines: Requires less overlap and allows faster flights.
Power lines: Detecting thin electrical cables against the sky is far more reliable with LiDAR.

Use photogrammetry when:

Open terrain: Quarries, mines, beaches, post-harvest agricultural land. Centimeter-level accuracy with much cheaper equipment.
Limited budget: A DJI Mavic 3 Enterprise with RTK (DJI has since launched the Matrice 4E in the same range) costs ~€4,200 and does excellent photogrammetry.
Need for texture and color: Native textured models, superior for presentations.
Building inspection: Facades and roofs with visually rich models.

Other options on the market in 2026

Autel EVO Max 4T: Note that this is not a LiDAR system. It combines a zoom camera, a wide-angle camera, a thermal camera and a laser rangefinder, which measures the distance to a single point but does not generate point clouds. It is built for inspection, not for LiDAR mapping.

YellowScan Mapper+: up to 240,000 shots/second, up to 3 returns, 4 cm accuracy according to the manufacturer, multi-platform compatible. Excellent for forestry and archaeology.

Livox Mid-360: A compact sensor (265 g) designed for robotics, with 200,000 points/second, a single return and a range of 40 m at 10% reflectivity (70 m at 80%). Adapting it to a drone means integrating GNSS/RTK positioning and software on your own, and the short range forces you to fly very low. It is fine for experimenting, not as an alternative to an aerial mapping system.

Processing software

DJI Terra: DJI’s software for processing L2 data. Good for simple deliverables, sufficient for most projects.

CloudCompare: Free and open-source. The interface won’t win design awards, but functionality is excellent.

Terrasolid: Industry standard. For demanding forestry and civil engineering, worth every cent.

Pix4Dmatic: From $125 per month, according to Pix4D’s price list. The best for combining photogrammetry + LiDAR in the same project.

Conclusion

LiDAR on drones is no longer elitist technology. With the Zenmuse L2 starting at around €10,500, precision 3D mapping is now within reach of many more operators. The key question is knowing when to use each technology — and in most cases, the answer is to use both.

If you already have a DJI Enterprise drone and do regular surveying, the L2 is the logical next step. The return on investment depends on the price charged per survey and the monthly volume of work.

For those wanting to try before committing, renting equipment or subcontracting the first surveys are good entry points. Precision 3D mapping is within everyone’s reach in 2026.

Frequently Asked Questions

Do I need special training to use LiDAR on drones?

For flying, you just need your regular drone license. For LiDAR processing, I strongly recommend specific training — not for sensor operation (which is automatic) but for post-processing and data interpretation. There are dedicated LiDAR processing courses that are worth the investment.

Does LiDAR work on cloudy or rainy days?

In fog or heavy rain, it loses range and accuracy significantly. On cloudy days without precipitation, it works perfectly — in fact, it works better than photogrammetry on cloudy days because it doesn’t depend on sunlight to measure distances. A frequently overlooked advantage.

What’s the real accuracy of drone LiDAR?

With the Zenmuse L2 and active RTK, DJI specifies 4 cm vertically and 5 cm horizontally at 150 m altitude. Ground control points (GCPs) measured on site let you confirm those values on each project.

Is the Livox Mid-360 worth it for mapping?

As an aerial mapping sensor, no. It has a single return (weak for forest) and a range of 40 m at 10% reflectivity, which forces you to fly very low, and it requires integrating GNSS/RTK positioning and software on your own. It is a good sensor for robotics and experiments, not a cheap shortcut to aerial LiDAR.

Miguel Santos

DroneFlyBlog

Keep exploring the world of drones

Get new buying, aerial photography and drone use guides to prepare for your next purchase or flight.