Close-Range Photogrammetry and Terrestrial Applications

What Is Close-Range Photogrammetry?

Close-range photogrammetry (CRP) applies the same principles as aerial photogrammetry but at distances typically under 300 meters — often just a few meters. You use overlapping photographs of a real-world object or structure, taken from multiple positions and angles, to reconstruct a precise 3-D model and extract measurements. Common field applications include documenting retaining walls, concrete structures, stockpile volumes, bridge abutments, and pavement distress. Because it requires only a calibrated camera and processing software, CRP is a cost-effective complement to total-station survey and drone mapping.

Planning the Camera-Station Network

A good camera network is the foundation of a successful CRP survey. Before going to the field, sketch the structure you are measuring and mark camera positions around it. Three principles guide the layout:

• Overlap: Each point on the surface should appear in at least three images, preferably five or more. Industry guidance recommends 60–80% overlap between adjacent images.
• Convergence angle: Camera axes should converge on the subject from multiple directions. Shots taken only straight-on (parallel axes) create weak geometry and depth errors. Aim for convergence angles of roughly 15–30° between adjacent stations.
• Coverage: Photograph the top, sides, and — where accessible and safe — the base of the feature. Gaps in coverage create holes in the mesh.

During planning, list every scale reference you will place in the field. At minimum, use two coded scale bars or three measured reference targets distributed around the subject. Scales placed at only one elevation or on only one face of a structure will not constrain the full model geometry.

Scale References and Coded Targets

Photogrammetric software like Agisoft Metashape can automatically detect coded targets — printed patterns that resemble QR codes. Each target has a unique ID so the software can match them across images without manual intervention. Scale bars are rigid bars with a precisely known length, typically printed or engraved with coded targets at each end. When you tell Metashape the exact distance between the two end targets, it scales the entire model to real-world units.

Before shooting, measure and record every reference distance with a steel or fiberglass tape. Steel tapes are preferred for highest accuracy; fiberglass tapes can stretch. Record measurements to the nearest millimeter in your field log.

Field Image Collection

Follow your planned network systematically. Keep the lens clean and the camera level or consistently tilted. Use a shutter speed fast enough to avoid motion blur — generally 1/250 s or faster in outdoor light. Avoid shooting into direct sunlight, which causes lens flare and overexposed highlights that confuse matching algorithms. Overcast days often produce the most uniform texture on walls and earthworks.

Work in pairs: one person photographs while the other records images, targets, and conditions in the field log. Log entries should include date, time, location, weather and light conditions, camera model and lens focal length, number of images captured, and any anomalies such as moving objects or wet surfaces.

Never position yourself under an unstable slope, beside a loaded retaining wall showing distress, or in a construction zone without required personal protective equipment. Wear a high-visibility vest whenever you are near vehicle traffic. Wear a hard hat if overhead hazards — falling materials, crane loads, or construction activity — are present. Inspect the structure and its surroundings before placing targets; report any apparent structural deficiency to the supervising instructor immediately and do not enter the hazard zone.

Processing in Metashape

Import your images into a new Metashape project. Add your coded targets and define the scale bar distances you measured in the field. The processing workflow follows these stages:

1. Align Photos — Metashape identifies common feature points and estimates camera positions.

2. Build Dense Cloud — A dense point cloud is generated from the aligned cameras.

3. Build Mesh — The point cloud is triangulated into a surface model.

4. Build Texture — Photographs are projected onto the mesh to create a photorealistic surface.

After processing, use Metashape's ruler and measurement tools to extract dimensions: lengths, widths, heights, and cross-sectional areas. Export the point cloud and mesh in formats useful for inspection deliverables (e.g., .LAS for point cloud, .OBJ for mesh).

Extracting Measurements and Checking Accuracy

For each dimension you measure in Metashape, also record the same dimension from your tape measure taken in the field. Enter both values into a comparison table. Calculate the discrepancy (absolute difference). A discrepancy within 2 cm is generally acceptable for construction inspection at this scale.

If a discrepancy exceeds 2 cm, investigate before concluding the survey failed. Common causes include: insufficient scale references, a scale bar that was not rigidly placed (it moved between images), camera-station geometry that was too narrow, or a tape measurement made to the wrong point on the surface. Document your explanation in the processing log — a well-reasoned explanation of a larger discrepancy demonstrates professional competence.

Connecting to Your Credentials

• FAA Part 107: CRP skills translate directly to drone-based inspection workflows. Understanding camera geometry, overlap, and scale reference placement is required whether your camera flies or is handheld.
• Esri ArcGIS Pro: Point clouds and meshes produced by CRP can be imported into ArcGIS Pro for integration with site plans, as-built drawings, and GIS datasets.
• NSPS CST: Surveyors increasingly use CRP for as-built documentation and deformation monitoring. CST technicians are expected to understand measurement accuracy, error sources, and proper field documentation.