DJI Zenmuse L3

With a laser beam divergence of just 0.25 mrad (1/e²), the spot size at the same range is about one-fifth that of Zenmuse L2, enabling detection of smaller objects such as power lines and branches. Higher energy density results in stronger penetration capability, significantly increasing ground point density beneath vegetation and enabling high-accuracy reconstruction of complex terrains and structures.

Supports Linear, Star-Shaped, and Non-Repetitive scanning modes for different operational scenarios:

• Linear: Achieves more consistent point cloud distribution, ideal for high-accuracy terrain mapping.
• Star-Shaped: Balances accuracy with multiple scan angles and penetration capability, well-suited for forests or dense urban environments.
• Non-Repetitive: Provides multiple scan angles and broader coverage, optimized for power line inspections and other complex structural surveys.

The dual RGB mapping cameras provide a horizontal field of view (FOV) of up to 107°, allowing each capture to cover a broader area. Even with a 20% LiDAR side overlap ratio, you can capture data for both digital orthophoto maps (DOM) and digital elevation models (DEM) in a single flight, thereby greatly improving operational efficiency.

The upgraded high-precision POS system achieves heading accuracy of 0.02° and pitch accuracy of 0.01° (post-processed, 1σ). The self-developed LiDAR delivers ranging repeatability of 5 mm at 150 m (1σ) and supports microsecond-level time synchronization. Each unit undergoes rigorous factory calibration to ensure consistently reliable survey results.

With dual 100MP RGB mapping cameras, Zenmuse L3 achieves a ground sample distance (GSD) of 3 cm even at a 300m flight altitude. Real-time position synchronization enables operation without ground control points (GCPs), ensuring high-quality imagery for precise mapping.

With a smaller laser spot and higher single-pulse energy, Zenmuse L3 offers superior penetration capability. In terrain mapping and similar applications, it captures more ground points with greater completeness, significantly improving the accuracy and integrity of terrain data for precise surface fitting and reconstruction.

When mounted on DJI Matrice 400, Zenmuse L3 delivers outstanding efficiency. At a flight altitude of 300 m, a single flight can cover up to 10 km², enabling up to 100 km² of daily coverage

DJI Pilot 2 automatically recommends flight parameters based on user-defined routes and generates a Density Profile Chart and point cloud distribution charts before takeoff, enabling users to preview point cloud results.

DJI Pilot 2 supports split-screen viewing of any two data sources—point cloud, RGB, or Vision Assist—allowing users to freely combine displays to suit their needs. During Power Line Follow or when the RGB cameras face downward, Vision Assist provides real-time forward-facing visuals, reducing flight risks and improving overall efficiency and safety.

Zenmuse L3 point cloud outputs are fully compatible with DJI FlightHub 2. Users can view and manage point cloud data in real time, easily generate detailed flight routes, annotate and share points of interest, and perform advanced line and area measurements directly in the cloud.

DJI Terra enables advanced Gaussian Splatting fusion-based reconstruction of point cloud and RGB data collected by Zenmuse L3, dramatically enhancing the realism and detail of 3D models. This capability broadens the application scope of LiDAR across diverse mapping scenarios and delivers higher-quality 3D model outputs.

DJI Modify supports intelligent semantic classification of point clouds, automatically distinguishing ground, vegetation, buildings, wires, electric towers, and more. After classification, users can efficiently adjust categories using Quick Selection and Structure Selection tools, further enhancing classification accuracy and processing efficiency.

When paired with DJI Modify, Zenmuse L3 delivers versatile outputs. DJI Modify automatically generates a wide range of mapping products from point cloud data—including semantic point clouds, DEM, TIN, point grids, and contour lines—and supports multi-layer visualization. These outputs meet the diverse needs of engineering and design CAD-related workflows.