Smartphones and photo analysis techniques are transforming field surveying! Traditional surveying required specialized equipment and a lot of effort, but recently, combining smartphones with SfM processing (photogrammetry) has made high-precision 3D measurement accessible to anyone. This article explains the benefits and latest trends of field surveying using SfM processing and introduces next-generation surveying methods that can be completed with just a smartphone. Using the solution called LRTK featured in the title as an example, we explore the concrete picture of how field surveying is being updated.

Challenges of field surveying and the need for digitalization

In construction and civil engineering, accurate measurement and recording of the shapes of built structures and terrain—so-called “field surveying” or “as-built management”—is indispensable. However, conventional surveying methods had several challenges. For example, they required surveyors who could operate specialized instruments such as total stations and levels, which increased personnel and time costs. On large sites, many survey points had to be measured one by one, causing a heavy workload. Because measurements were taken point by point, it was also difficult to grasp the overall picture. Differences from the design in areas other than measured points could be overlooked, and problems sometimes emerged later.

To address these issues, on the back of initiatives like the Ministry of Land, Infrastructure, Transport and Tourism’s *i-Construction*, digitalization of field sites (field DX) is progressing. One rapidly spreading approach is the use of 3D measurement technologies. Methods that acquire the entire site’s shape as point cloud data at once—using laser scanners or drone photogrammetry—and allow planar and volumetric understanding are becoming the new norm. However, the introduction of expensive 3D laser scanners posed cost and expertise barriers. Attention has therefore turned to the more accessible combination of SfM photogrammetry and smartphones.

New possibilities opened by SfM processing for field surveying

SfM processing (Structure from Motion photo analysis) is a technique for reconstructing 3D shapes from multiple photos. For example, if you photograph a site with a smartphone or digital camera and process those photos with dedicated software, a 3D model of the entire site (point cloud or mesh) is generated. This enables obtaining a high-density digital point cloud that covers the entire site, rather than a collection of manually acquired points. With SfM-based photogrammetry, you can record large areas in a short time and later flexibly measure dimensions at arbitrary points or extract cross-sections for analysis.

Moreover, SfM photogrammetry can be practiced with common equipment. If you have a smartphone with a high-resolution camera, you can scan a site without special surveying instruments. While drone-based SfM from aerial photos was common, walking around on foot and taking photos with a smartphone is increasingly sufficient to collect the necessary data. Because smartphones are easy for anyone to use, a major advantage is that site personnel themselves can perform measurements without specialized operators.

Why easy, high-precision 3D measurement with a smartphone became possible

The combination of smartphones and SfM processing enabling easy and high-precision 3D measurement is due to technological advances. First, recent smartphones have dramatically improved camera performance and can capture very high-resolution photos, which directly improves SfM accuracy. In addition, many phones now include IMUs and barometers, and some models support high-precision GNSS, making it easier to utilize position and orientation information in analysis.

The improvement of computer performance for processing large numbers of images and the expansion of cloud services are also key. If you upload photos taken with your smartphone to the cloud, services can automatically perform heavy SfM processing on servers. This allows you to obtain analysis results on-site without carrying a specialized PC.

A decisive development was the emergence of smartphone solutions combined with RTK-GNSS (real-time kinematic GPS). RTK enables obtaining photo capture locations and coordinates of feature points on the model with centimeter-level (half-inch-level) position accuracy, so SfM-generated 3D models can be given an accurate scale and absolute coordinates. Previously, photogrammetry required measuring reference distances or placing ground control points to scale a model to real-world dimensions. RTK-enabled smartphone measurement drastically reduces such steps, achieving both ease of use and improved accuracy.

Field use of smartphone surveying made possible by LRTK

Against the backdrop of these technological advances, “LRTK”, sometimes called an all-purpose surveying app, has appeared. LRTK is a system for attaching an ultra-compact RTK-GNSS receiver to a smartphone (iPhone/iPad) and is provided for field use by the developer, Refixia Co. With LRTK, a single worker can perform surveying, point-cloud measurement, and photography on-site using just a smartphone—an all-in-one workflow. Tasks that previously required separate surveying and photo-record teams can be completed with one device using LRTK, leading to significant efficiency gains.

Particularly noteworthy is the photogrammetry feature built into the latest LRTK. This implements SfM processing on the smartphone itself, enabling instant 3D model generation from photos taken on-site. LRTK integrates high-precision RTK location information with SfM algorithms, so each point in the generated 3D model is assigned absolute coordinates (latitude, longitude, elevation) from global positioning. Therefore, by viewing the completed model you can digitally measure dimensions, slopes, areas, and volumes as if you were on-site. Because the model already aligns with a geodetic coordinate system, it is also easy to overlay the resulting point cloud or 3D model onto BIM/CIM design data or GIS maps for downstream use.

LRTK’s photogrammetry is reported to have an accuracy of about ±5 cm (±2.0 in), achieving a level of absolute accuracy that conventional photogrammetry found difficult. Moreover, because this processing can be completed on the smartphone itself, models can be generated on-site even in mountainous areas or other locations with unstable internet connectivity. Eliminating the time lag of returning to the office for PC analysis makes immediate on-site verification possible—a revolutionary improvement.

Benefits of combining SfM processing with RTK positioning

Let’s summarize the advantages of combining SfM processing with RTK positioning for field surveying.

• Greatly expanded measurement coverage: Instead of point-by-point surveying, you can acquire the overall shape of the site, allowing later analysis of arbitrary locations. This reduces oversights and the number of return visits to the site for additional measurements.

• Improved efficiency and reduced costs: With a single smartphone, one person can perform surveying, cutting labor and equipment costs. Less time is spent on movement and setup, enabling wide-area coverage in a short time.

• High-precision data acquisition: Using RTK-GNSS gives acquired data absolute coordinates, so the resulting 3D models have accurate scale and conform to public reference coordinate systems. This eases comparisons with design drawings and other survey data in downstream processes.

• Improved safety: Dangerous slopes and high areas can be photographed remotely, reducing the need for workers to enter hazardous zones, which contributes to overall site safety.

• Accumulation of digital records: If point clouds and 3D models are stored in the cloud, the site state can be reviewed later. This makes it easier to share and verify construction progress records and evidence of completed work among stakeholders.

In this way, smartphone surveying that fuses SfM processing with RTK positioning is an innovative approach that brings benefits for quality, efficiency, and safety.

Conclusion: Start next-generation field surveying with LRTK

Photogrammetry SfM, smartphones, and RTK positioning—this combination is updating field surveying right now. Surveying, once the domain of specialists, is becoming a digital tool accessible to anyone on-site. Solutions like LRTK are representative examples and are already attracting attention as tools that accelerate field DX.

If you feel limitations or inefficiencies in traditional surveying methods, consider simple surveying with LRTK. This system, which enables high-precision 3D measurement with a smartphone and familiar devices, is designed to be approachable even for those trying digital photogrammetry for the first time. By adopting next-generation field surveying that leverages the latest technologies, you can achieve both efficiency gains and accuracy improvements. Bring the wave of DX to your site with LRTK and SfM processing, the new normal in field surveying.

SfM processing introduction guide for beginners: Complete support from preparation to 3D model creation

What is SfM processing? What preparations are needed, and how do you create a 3D model from photos? This article thoroughly explains the basic concepts of SfM (Structure from Motion), common pitfalls for beginners, required equipment and preparation, photography tips, and the actual processing steps. Even those attempting SfM processing for the first time should be able to understand the flow from preparation to completed 3D model after reading this.

What is SfM processing? A technique to generate 3D models from photos

SfM processing is a technique for reconstructing a target or site’s three-dimensional model from multiple photographic images. SfM stands for “Structure from Motion,” meaning it estimates structure from camera motion. By taking many photos from various angles with a drone, digital camera, or smartphone, and analyzing the displacement of feature points between images, the three-dimensional shape of the subject is computed. The resulting 3D data can be a point cloud or polygon mesh, and it is possible to generate texture-mapped 3D models with photo color projected onto the surface, as well as orthophotos that look like overhead views. In short, with SfM processing, you can create precise 3D models of real objects using only a handheld camera, without a specialized 3D scanner.

Equipment and preparation needed for SfM processing

For beginners starting SfM processing, first check the necessary equipment and preparatory steps. Basically, expensive specialized equipment is not required, and you can start with common items.

• Camera or smartphone: Prepare a device capable of taking high-quality photos. Ordinary digital cameras or modern smartphones are usually sufficient. Cameras with higher pixel counts and better lens performance will capture finer detail.

• Computer: You need a PC to process the captured photos. SfM processing analyzes a large number of images, so a fairly capable PC (with adequate memory and GPU) is desirable, but if the number of photos is small, a standard-spec computer can suffice.

• SfM software: Prepare software dedicated to generating 3D models from photos. There are various commercial and open-source SfM packages; for beginners choose one with an easy interface and comprehensive tutorials. Some software provides a GUI for easy processing, while others allow fine-tuning of many settings.

• Target markers and scale (as needed): If you want the model to reflect real-world scale, prepare a scale bar (ruler) or fiducial markers to include in the photos. For example, place a 1 m (3.3 ft) rod or a checkerboard near the subject and capture it; during post-processing you can use that known distance to scale the model. This is not essential, but for measurement purposes where accurate dimensions are required, it’s useful to prepare.

After assembling this equipment, decide on the subject (site or object) and plan your on-site photography. The next section covers key points for taking photos suitable for SfM.

Photography tips: overlap and multiple angles are key

The success of SfM processing depends heavily on the quality of the photos. Beginners should keep the following points in mind when shooting.

• Sufficient number of photos and overlap: To obtain a high-quality 3D model, take many photos covering the entire subject, ensuring 60–80% overlap between consecutive images. If overlap is low, the software may not find correspondences between images and the model may not reconstruct well. It’s safer to err on the side of taking too many photos than too few.

• Shoot from various angles: Capturing the same subject from different angles and heights provides three-dimensional information. Avoid only top-down or side views; shoot diagonally and surround the subject. For large objects like buildings or terrain, plan shooting positions to capture the subject from all around.

• Watch focus and blur: All photos should be in focus on the subject. Blurry or out-of-focus images hinder analysis, so use a tripod, set a fast shutter speed, or use smartphone stabilization features. In low-light situations like indoors or at dusk, avoid raising ISO too high (which increases noise); instead, add lighting to capture sharp images.

• Keep exposure and settings consistent: If photos with drastically different brightness or color are mixed, the software may struggle to recognize the same subject. Try to fix camera settings (exposure, white balance) and maintain a consistent image set. If using auto mode, correct settings if they change during shooting.

• Get close to the subject: Rather than zooming from a long distance, get as close as possible to capture fine details. For very large structures, include some wide shots to cover the overall shape along with close-ups for detail.

Following these points will yield image data that SfM software can process effectively. Once shooting is complete, proceed to SfM processing in the software.

Workflow for generating 3D models with SfM software

Post-photography processing is handled by software, but understanding the general workflow helps. A typical SfM processing sequence is as follows.

• Load photo data: Import the multiple photos into the SfM software. Camera parameters (lens focal length, sensor size, etc.) may be automatically detected, but confirm settings if necessary.

• Feature detection and matching: The software detects feature points within each image (high-contrast patterns, corners, etc.) that serve as alignment cues. It then matches the same points (correspondences) across multiple images. The more common feature points found across image pairs, the higher the subsequent computation accuracy.

• Camera pose estimation (photo alignment): Using the matched features, the software computes where and how each photo was taken—estimating external camera parameters (position and orientation)—and simultaneously computes 3D coordinates of the feature points (a sparse point cloud). In simple terms, this step reconstructs camera viewpoints and the coarse shape of the subject as points.

• Dense point cloud generation: Based on estimated camera poses and the sparse point cloud, the software performs pixel-level analysis to create a detailed point cloud. This produces a high-density set of points that accurately represents features of the subject, such as building details and terrain undulations.

• 3D mesh and texture creation: If needed, generate a polygon 3D mesh from the point cloud and apply texture mapping using photographic color information to produce a realistic-looking 3D model. For ground-surface captures, some software can also output an orthomosaic (a nadir-view photo map).

• Scale adjustment and coordinate alignment: Align the model to real-world dimensions and coordinate systems. If a scale bar was included in the photos, use its known distance to scale the entire model. To georeference the model to survey coordinates, measure a few points on-site and assign their coordinates to corresponding model points (a process called georeferencing).

• Export and utilization: Export the completed 3D model or point cloud in formats needed for downstream use. In construction, point clouds may be imported into CAD software for volume calculations; in cultural heritage, 3D models may be published on the web, and so on.

Terminology and steps vary slightly by software, but the overall flow is as above. Beginners can rely on automatic processing at first, and as they gain experience, they can fine-tune parameters in each step for higher accuracy.

Tips for obtaining high-precision models

To achieve satisfying 3D models even on your first attempts, keep several accuracy-improvement tips in mind.

• Use high-resolution, low-distortion images: Image quality directly affects model accuracy. Shoot at the highest feasible resolution and, if using wide-angle lenses, perform distortion correction or camera calibration beforehand to compensate for lens characteristics.

• Ensure sufficient overlap and multi-view coverage: As noted earlier, higher overlap and multiple viewing angles allow the software to match feature points with confidence. The mantra for SfM success is “take many overlapping photos from various angles.”

• Choose subjects with adequate features: SfM tracks surface features, so uniform surfaces like plain white walls or glass without texture are problematic. Where possible, select subjects with texture or apply markers to feature-poor areas to create artificial cues.

• Balance data volume and processing time: Excessive photo counts can greatly increase processing time or cause PC instability. Start with a few dozen images and gradually increase numbers as needed. Some software offers optimizations (e.g., temporary image downscaling for feature detection) that reduce computation without sacrificing final quality.

• Consider partial retakes: If part of the point cloud is sparse or distorted, it may indicate insufficient photos for that area. If possible, retake additional shots of the problematic region and reprocess to improve quality.

Applying these practices will raise the accuracy and completeness of SfM-generated models.

Use cases for SfM: from simple applications to professional surveying

SfM-generated 3D models are used across many fields. Here are examples from beginner-friendly to professional applications.

• Construction and civil engineering surveying: Capture construction sites with drones or smartphones to create 3D models of terrain and structures for as-built management and volume calculations. SfM enables obtaining a planar depiction of the site at once, streamlining tasks that previously required many point measurements.

• Cultural heritage and archaeological documentation: Projects worldwide use SfM to create detailed digital archives of historical buildings and ruins. Photogrammetry allows non-contact 3D capture of valuable artifacts from photographs.

• Accident site analysis: SfM reconstruction of traffic accidents or disaster sites supports post-event analysis and reporting. Situations that are hard to convey in flat photos can be intuitively understood in 3D, aiding investigation.

• Game and VR content creation: Scan real-world objects (sculptures, props) with SfM to import realistic 3D assets into games and VR. This is often faster than modeling from scratch and is gaining traction in creative industries.

• Education and research: SfM is used in academic surveys for terrain and vegetation mapping, and architecture students use SfM to digitize models for design studies. Its use in education and research continues to expand.

As shown, SfM is versatile and useful from hobbyist experiments to professional applications.

Precautions to avoid failure

Finally, here are some precautions to ensure smooth SfM success and reduce failure risk.

• Clear the surroundings before shooting: Moving people or vehicles in the scene can change features between photos and disrupt feature matching. Shoot in as static an environment as possible and remove obstructive items beforehand.

• Avoid mixing unrelated photos: Use only photos of the same subject for processing. Mixing images from different locations can prevent correct reconstruction. Organize photos into folders after shooting.

• Check intermediate results: Monitor alignment results (camera positions and sparse point cloud). If a camera position looks clearly wrong, consider excluding that photo or retaking it. Early fixes save rework later.

• Prepare for software crashes: For long processing jobs, save progress frequently if the software allows. This prevents losing an overnight run to an unexpected error.

• Observe laws and privacy: When using a drone, comply with aviation regulations; avoid photographing confidential materials; and respect privacy when people may be captured. Confirm legal and ethical constraints for your purpose and location.

Following these points during preparation, shooting, and processing will reduce the chance of major problems. You may need trial and error at first, but experience will help you develop reliable techniques.

Conclusion: make better use of photos with SfM processing

SfM processing, which creates 3D models from photos, is an attractive technology that beginners can implement by following proper steps. Without special sensors, you can obtain high-precision 3D data using just a camera and a PC, and the range of applications will continue to expand. Start with small targets and give SfM a try.

If you want to apply SfM-generated 3D models to actual field measurement or work, consider simple surveying with LRTK. LRTK combines a smartphone with high-precision GNSS, enabling seamless photogrammetry and positional measurement. It is beginner-friendly and can make SfM-based 3D measurement easier and more reliable. Use the latest technologies to fully unlock the potential of your photos.