AP60 Vision RTK: Fill Drone Survey Gaps with Ground-Level Visual Measurement
DJI drones capture excellent aerial photogrammetry across open terrain but cannot cover building interiors, areas under canopy, recesses beneath overhangs, dense urban street corridors, or any feature blocked from above. The APEKS AP60 Vision and AP80 Pro use front-facing cameras to capture ground-level video sweeps of these areas — the receiver automatically derives 3D absolute coordinates from stereo photo pairs using the RTK position and IMU angle. No additional drone flight required. The ground-level visual measurement data exports in standard formats and merges directly with drone photogrammetry outputs in Pix4D, Agisoft Metashape, or DJI Terra to produce a complete 3D model of the site.
Drone survey has transformed large-area topographic mapping. A single DJI Matrice 350 RTK flight can capture a 50-hectare site in under an hour. But every drone survey has the same structural limitation: the camera points down. Features that cannot be seen from above — building façades, areas under tree canopy, under-eave voids, confined urban corridors, pipe culvert interiors, cliff faces — produce gaps in the photogrammetric model that no amount of additional flight altitude or overlap can fill. The APEKS AP60 Vision and AP80 Pro address this directly: ground-level visual measurement captures the areas the drone missed, in the same coordinate system, in the same field session. This guide covers how the workflow operates and where it delivers the most value.
The Drone Survey Coverage Problem
Drone photogrammetry generates accurate 3D models of surfaces visible from above. For flat or gently undulating open terrain, coverage is nearly complete. Coverage gaps appear when:
VERTICAL SURFACES:
Building façades, retaining walls, bridge abutments, embankment faces, and cliff walls are perpendicular to the drone camera's downward view. Oblique drone imagery captures partial coverage but introduces geometric distortion and scale errors that reduce accuracy on vertical surfaces to typically ±50–200mm — insufficient for as-built documentation, heritage recording, or structural assessment.
AREAS UNDER CANOPY:
Dense tree canopy, crop rows, and forest areas block the drone's view of the ground surface beneath. Ground topography under canopy is absent or poorly defined in the photogrammetric model.
CONFINED AND ENCLOSED AREAS:
Urban street corridors with tall buildings on both sides, pipe culvert interiors, drainage channel bases, under-eave recesses, and tunnel portals cannot be safely flown or adequately captured by a downward-facing camera.
OVERHANGING FEATURES:
Bridge deck undersides, elevated structures, overhanging rock faces, and cantilevered building elements are invisible from above regardless of flight altitude.
The result: every drone survey of a complex site produces a model with gaps at exactly the features that matter most for engineering, heritage, and infrastructure work.
How RTK Visual Measurement Works
Visual measurement on the AP60 Vision and AP80 Pro uses the front-facing camera to capture a continuous video sweep of the target surface. As the operator moves the receiver across the face of a building, wall, or structure, the receiver automatically extracts stereo photo pairs from the video stream.
THE CALCULATION:
For each stereo pair, the receiver knows:
- The RTK antenna position at each frame (from UM980 GNSS)
- The camera orientation at each frame (from 120° IMU)
- The baseline between stereo frames (from receiver movement)
From these inputs, ApekSurv derives 3D absolute coordinates for points across the captured surface — the same photogrammetric principle used in drone mapping, applied at ground level with RTK-grade absolute positioning.
OUTPUT:
The visual measurement output is a set of 3D coordinates referenced to the project datum — the same coordinate system used by the drone survey. The data exports in standard formats (CSV, DXF, compatible with Pix4D, Metashape, and DJI Terra).
NO PRISM, NO TARGETS:
Visual measurement requires no reflective targets, no prism placement on the surface, and no physical contact with the structure. The operator walks the face of the feature at a practical working distance and the receiver does the rest.
IMPORTANT DISTINCTION:
Visual measurement derives discrete 3D coordinates from stereo pairs — it is not a LiDAR or SLAM system. For dense surface modelling of complex geometry, it complements rather than replaces LiDAR scanning. For standard engineering survey accuracy on building façades, walls, and structures, it covers the requirement without additional equipment.
What AP60 Vision Adds to a Drone Survey Workflow
The AP60 Vision is APEKS's dedicated visual measurement receiver. In the context of a drone survey workflow, it fills the coverage gaps left by aerial photogrammetry.
BUILDING FAÇADES AND VERTICAL SURFACES:
Walk the face of each building elevation while the drone covers the roof and surrounding ground. The AP60 Vision captures the façade at ground level with RTK-grade absolute positioning. The two datasets share the same coordinate system and merge in post-processing.
UNDER-CANOPY GROUND SURFACE:
Where drone imagery cannot penetrate tree cover, the AP60 Vision captures ground topography by visual sweep at ground level. Combined with the aerial model, the merged output provides a complete terrain model including beneath-canopy areas.
3D MODELLING OUTPUT:
The AP60 Vision supports 3D modelling export — the visual measurement data can be imported directly into 3D modelling software alongside the drone photogrammetry output. For heritage recording, as-built documentation, and urban modelling, the combined dataset provides the complete site model that neither source alone could deliver.
AR STAKEOUT ON THE SAME INSTRUMENT:
The AP60 Vision also carries a bottom-facing AR stakeout camera — the same instrument that captures building façades in the morning can set out construction points from the AR overlay in the afternoon. No instrument change required.
AP60 Vision vs AP80 Pro — Which for Drone Complement Work
Both the AP60 Vision and AP80 Pro support visual measurement and 3D modelling for drone complement work. The AP80 Pro adds a 120m green laser for features beyond camera range. The choice depends on project complexity.
| Feature | AP60 Vision | AP80 Pro |
|---|---|---|
| Visual Measurement | ✓ | ✓ |
| 3D Modelling Output | ✓ | ✓ |
| AR Stakeout (bottom camera) | ✓ | ✓ |
| Green Laser (120m) | ✗ | ✓ |
| Laser Offset Measurement | ✗ | ✓ |
| GNSS Battle 2026 | — | 🥇 Grand Champion |
| Best for drone complement | Standard façade and canopy work | Complex sites needing laser + visual in one session |
For standard drone complement work — building façades, under-canopy ground, urban corridors — the AP60 Vision covers the requirement at lower cost. For sites that combine visual measurement of surfaces with laser measurement of inaccessible features (bridge abutments, overhead structures, features across live traffic), the AP80 Pro handles both in a single instrument.
Step-by-Step: Ground Complement Workflow
The Core Problems This Solves
Symptom: The drone photogrammetry model of an urban site shows accurate rooftop and ground surface geometry but the building walls appear as blurred, distorted surfaces with ±100–200mm accuracy. The as-built documentation cannot be submitted for planning or heritage purposes because façade accuracy is insufficient.
Cause: Drone cameras point downward. Oblique imagery at standard mapping altitude (80–120m) captures building walls at a shallow angle that introduces significant geometric distortion. Higher oblique flight angles improve wall coverage but reduce ground accuracy. No single drone flight configuration optimises both simultaneously.
Fix: After the drone flight, walk the AP60 Vision along each building elevation at 3–8m working distance. Visual measurement captures each façade at perpendicular incidence with RTK-grade absolute positioning. The façade data merges with the drone roof and ground model in post-processing, producing a complete building model with consistent accuracy across all surfaces.
Symptom: The drone model of a forested or heavily vegetated site shows accurate canopy surface but the ground beneath the trees is absent or interpolated. Drainage channels, pathways, property boundaries, and infrastructure beneath the canopy are missing from the survey output.
Cause: Dense canopy blocks the drone camera's view of the ground surface beneath. Even with ground filtering algorithms, photogrammetric models cannot reconstruct surface geometry that was never imaged.
Fix: Walk the AP60 Vision through the under-canopy zones, capturing ground topography by visual sweep at ground level. In open forest where sky view is sufficient for RTK Fixed solution, the visual measurement provides ground coordinate data that the drone model cannot produce. The combined dataset covers both canopy surface (drone) and ground surface (AP60 Vision) in a single coordinate system.
Symptom: The drone photogrammetry data and the ground survey data were collected using different correction sources. When imported together in Metashape or Pix4D, the two datasets show a systematic offset — the ground survey points do not align with the drone model surface. Manual registration is required, introducing additional uncertainty.
Cause: When the drone survey and the ground survey use different correction sources — different CORS mountpoints, different base station coordinates, or different days with different base setups — small systematic offsets accumulate between the two datasets.
Fix: Use the same correction source for both the drone survey and the AP60 Vision ground session. If using a local base station, use the same AP10, AP20, or MAX5 on the same control point for both sessions. If using CORS, use the same mountpoint. This ensures both datasets reference the same absolute coordinate — they align in post-processing without manual registration.
Field Deployment Scenarios
SCENARIO 1 — URBAN HERITAGE BUILDING DOCUMENTATION:
DJI Matrice 350 RTK captures rooftop and surrounding ground model. AP60 Vision captures all four building elevations by ground-level visual sweep at 3–5m working distance. MAX5 base station provides corrections to both drone and AP60 Vision from the same control point. Combined dataset produces a complete building model for heritage record submission — rooftop, ground, and all four façades in a single coordinate system.
SCENARIO 2 — INFRASTRUCTURE CORRIDOR WITH MIXED TERRAIN:
Drone covers the open corridor — road alignment, earthworks, open ground. AP60 Vision covers drainage channel bases, culvert interiors, and retaining wall faces along the same corridor that the drone cannot reach. AP40 Laser+ measures bridge abutment positions across water features by laser offset. All three datasets reference the same MAX5 base — they merge without registration in post-processing.
SCENARIO 3 — FORESTED CADASTRAL BOUNDARY SURVEY:
Drone captures canopy surface and open clearings. AP60 Vision carried through the forest boundary zone captures ground topography and boundary feature positions beneath the canopy. RTK Fixed maintained in open forest conditions with 1408-channel full-constellation tracking. Combined output provides the complete property model required for cadastral submission.
FAQ
Can the AP60 Vision replace a drone for mapping?
No — and it is not designed to. The AP60 Vision covers ground-level surfaces and confined areas that drones cannot reach. For large open areas, drone photogrammetry remains faster and more comprehensive. The AP60 Vision's role is to fill the structural gaps in drone coverage — vertical surfaces, under-canopy ground, enclosed features — not to replace the aerial survey component. The two methods are complementary, not competitive.
What software can I use to merge AP60 Vision data with drone photogrammetry?
ApekSurv exports visual measurement data in formats compatible with the major photogrammetry platforms: Pix4D, Agisoft Metashape, and DJI Terra. Because both the drone data and the AP60 Vision data share the same RTK coordinate system, they align directly on import without manual GCP matching. Confirm the export format with your specific software version before the field session.
How far from the surface should I walk when capturing visual measurement?
For standard building façade and wall capture, a working distance of 3–8m gives good stereo geometry. Closer distances (1–3m) improve detail for complex surfaces. At distances beyond 15m, stereo pair quality and coordinate density reduce. Walk at a steady pace of approximately 0.5–1m per second for consistent frame extraction. Avoid rapid direction changes or stops that introduce motion blur in the stereo pairs.
Does visual measurement work at night or in low light?
Visual measurement requires ambient light for the camera to capture usable imagery — it does not work in darkness. In low light (dawn, dusk, or heavily overcast conditions), image quality degrades and stereo pair extraction becomes less reliable. For optimal results, capture in even natural light — overcast conditions are often better than direct sunlight, which creates high-contrast shadows on vertical surfaces that reduce feature matching quality.
FILL THE GAPS YOUR DRONE CANNOT REACH.
The AP60 Vision and AP80 Pro capture building façades, under-canopy ground, and confined features at ground level — in the same coordinate system as your drone survey. One field session. Complete 3D model. No additional flight required.
Send an Inquiry → WhatsApp Us →References
- ISO 17123-8:2015 — Field Procedures for GNSS RTK
- APEKS AP60 Vision Technical Datasheet, 2026
- APEKS AP80 Pro Technical Datasheet, 2026
- ApekSurv Field Software User Guide, 2026
- Unicore Communications UM980 Product Brief