Comprehensive Guide to Photogrammetry Technology: How Smartphone RTK Enables Centimeter-Level Accuracy 3D Surveying (centimeter-level accuracy (half-inch accuracy))

In recent years, the importance of three-dimensional (3D) surveying has been growing in construction and civil engineering. For example, using 3D models allows a three-dimensional understanding of site conditions, improving efficiency in planning, construction management, and maintenance. Against this backdrop, the technology of photogrammetry—generating precise 3D models from photographs—has attracted attention.

Photogrammetry is a method that analyzes multiple photographic images taken by drones or cameras to create point cloud data and 3D models. While it efficiently digitizes site geometry, traditional approaches posed several challenges in ensuring accuracy and in data processing. In particular, aligning the produced model to an accurate coordinate system required the installation of ground control points (GCPs) and additional surveying work, making it difficult to obtain immediate results.

A recently notable approach combines RTK (Real-Time Kinematic) technology with smartphones. By equipping a smartphone with a high-precision GNSS receiver, each captured photo can be tagged with centimeter-level position coordinates (half-inch accuracy), enabling high-precision 3D surveying easily without large specialized equipment.

This article explains the basic principles of photogrammetry, the traditional challenges and how fusing it with RTK improves accuracy, and the benefits of using smartphone RTK. It also covers how to utilize point cloud data obtained by photogrammetry and presents concrete application cases in infrastructure, construction management, and inspection. At the end, the article introduces a new smartphone-based surveying solution called “LRTK” and touches on how to start simple yet highly accurate 3D surveying.

How Photogrammetry Works and Its Characteristics

Photogrammetry refers to the technique of using photographic images to measure the shape and dimensions of objects and to create 3D models. For example, by photographing buildings or terrain from various angles with a digital camera or smartphone and analyzing those multiple images with specialized software, the subject’s three-dimensional shape can be reconstructed with high accuracy. The software detects many feature points (identical locations that serve as landmarks) appearing in each photo and, based on their correspondences, calculates the spatial coordinates of each point using the principle of triangulation.

This computation process is supported by image-processing techniques called Structure from Motion (SfM) and Multi-View Stereo (MVS). Simply put, by finding common points across multiple photos and resolving the geometric relationships between camera positions and the subject, the 3D positions of points are determined. The resulting large set of points (point cloud) includes X, Y, Z coordinates for each point plus color (RGB) information derived from the photos. Therefore, when the point cloud is visualized, it becomes a realistic 3D model that reflects the object’s shape and surface appearance.

A key feature of photogrammetry is that it can be used with ordinary digital or smartphone cameras, making it a low-cost way to record wide areas. There is no need to physically touch the object or prepare a special scanner; once photos of the site are taken, detailed 3D data can be obtained afterward. The density (resolution) of the resulting point cloud depends on the image resolution and shooting distance, but under favorable conditions, high-density point clouds capable of capturing millimeter-level (millimeter-level (0.04 in)) detail can be generated.

On the other hand, photogrammetry struggles to reproduce surfaces accurately when there are insufficient features. For example, a uniformly colored, patternless wall or reflective/transparent surfaces like glass or water make it difficult to detect feature points on photos, causing point cloud gaps. In such cases, artificial markers can be applied to provide features, or photogrammetry can be combined with other measurement methods to supplement missing data. Even so, because photogrammetry can quickly digitize an entire object, it significantly contributes to improving the efficiency of on-site measurement.

Challenges in Traditional Photogrammetry

Although 3D models and point clouds generated by photogrammetry have high relative shape accuracy within the subject, they may lack accuracy in absolute coordinates (real-world geodetic coordinate systems). Traditionally, to solve this problem, multiple known coordinate points (ground control points: GCPs) must be installed on site, photos must be taken so those points appear, and the model must be aligned to those points during post-processing. GCPs must be measured precisely with total stations or GNSS surveying equipment, and the wider the survey area, the more points and measurements are required. This increases on-site labor and time, detracting from the convenience of photogrammetry.

Photo surveying also demands considerable computational resources and time for image processing. When using hundreds of high-resolution photos, it is not uncommon for dedicated PCs to take hours to days to process them. Historically, workflows relied on cloud services or in-house high-performance servers to process large numbers of photos and then download the point cloud afterward. Even if one wanted to check a 3D model on site immediately, processing wait times reduced real-time capability.

Another accuracy-related issue is that conventional consumer cameras and smartphone GPS typically have errors of several meters (several ft), so generating a model using only the coordinates recorded with photos can result in scale and position discrepancies compared to reality. Ultimately, GCP correction and scale adjustments in post-processing were necessary, making such data difficult to use directly for surveying applications that require centimeter-level accuracy (half-inch accuracy).

Accuracy Improvements by Integrating RTK

The trump card to solve these issues is the integration of RTK (Real Time Kinematic) with photogrammetry. RTK is a positioning technique that dramatically improves GNSS accuracy by having both a base station installed at a known position and a rover receive satellite signals simultaneously; the base station sends real-time correction information to the rover, which corrects errors and achieves positioning accuracy within a few centimeters. Ordinary GPS positioning has errors of several meters (several ft), but applying RTK corrections can improve positioning accuracy to the centimeter level.

Combining RTK with photogrammetry provides the advantage of giving acquired 3D models and point clouds absolute accuracy. Specifically, if camera positions are recorded precisely with RTK-GNSS at the time of photography, each photo’s location (latitude, longitude, height) will be accurate. Photogrammetry software can use that information to reconstruct the model in a geographic coordinate system, so the resulting 3D model is displayed from the start at the correct position and scale in the public coordinate system (for example, the national geodetic reference frame). This greatly reduces the need for GCP-based alignment tasks and significantly lowers the model’s absolute error.

For example, in aerial photogrammetry using drones, an RTK-unaware aircraft might need more than 10 GCPs, whereas an RTK-equipped drone can drastically reduce the number of GCPs required. In some cases, reports indicate that a point cloud produced without any GCPs can still match map coordinates within 5 cm (2.0 in). In short, RTK dramatically improves photogrammetric positioning accuracy, allowing on-site 3D data to be integrated immediately into design coordinate systems.

Moreover, leveraging RTK brings photogrammetry closer to solving the previously problematic real-time processing on site. Photos tagged with RTK-GNSS have accurate initial positions, which helps the photo alignment computation (bundle adjustment) converge faster and more stably. Because coordinates are already aligned, rapid processing can be performed on-site using laptops or mobile devices, and in some cases, point cloud generation and simple analysis can be completed right there. The fusion with RTK has led to a leap forward in both accuracy and immediacy for photogrammetry.

New Developments in Surveying with Smartphone RTK

As the benefits of combining RTK and photogrammetry become clearer, smartphone RTK has attracted particular attention in recent years. Smartphones are devices with high-performance cameras and computing power; by combining them with a compact RTK-GNSS receiver, each person’s smartphone can function as a high-precision surveying instrument.

Traditionally, RTK surveying required dedicated expensive GNSS receivers and bulky gear, and operation demanded specialist knowledge. Recent technological advances have produced palm-sized RTK receivers with integrated antennas that can be attached to a smartphone or tablet to enable centimeter-level positioning. Because smartphones are typically internet-connected, they can access national reference networks or commercial correction services via Ntrip to receive RTK correction data in real time. This means high-precision positioning is possible nationwide without carrying dedicated radio devices or base stations, as long as you have a smartphone.

The main advantages of smartphone RTK are its ease of use and immediacy. There is no need for elaborate equipment setup upon arrival at a site; launching an app and tapping a few buttons on the phone screen is enough to start positioning. Photogrammetric measurements by photograph can be performed on the same device, enabling the entire workflow from image capture to point cloud generation to be completed on-site. You can immediately check 3D models on the smartphone, measure required dimensions, and share data with team members.

Additionally, the smartphone UX (user experience) is a significant benefit. Intuitive touch operations and familiar interfaces make the technology accessible even to non-specialist technicians. Tasks that previously required extensively trained surveyors can now be guided by smartphone apps so that general technicians can conduct data acquisition and processing themselves. In other words, smartphone RTK is democratizing high-precision surveying and promoting diverse on-site use.

Key advantages of introducing smartphone RTK include:

• Mobility: Only a smartphone and a small receiver are required, making equipment transport easy. Surveying can be done nimbly even in mountainous or confined sites.

• Simple Operation: Intuitive app-based setup and measurement reduce the need for specialist knowledge and lower training costs.

• Real-Time: Position correction and data processing can be performed on site, enabling immediate verification and sharing of results.

• Low Cost: Initial investment is lower than with traditional high-priced surveying gear, making distribution to multiple personnel easier.

• High Versatility: Since most people are already familiar with smartphones, adoption is easier and applications are broader.

Utilizing Point Cloud Data from Photogrammetry

Point cloud data and 3D models generated by photogrammetry can serve as digital twins of a site and have many uses. A point cloud is a collection of innumerable measurement points, so you can measure distances between arbitrary points on the model or calculate areas and volumes. For example, calculating earthwork volumes from point clouds obtained at an excavation site allows quick, accurate assessment of excavation volumes; measuring dimensions on a 3D model of a structure enables drawing creation without physically scaling on site.

Point cloud data also include color information at each point, making it easy to visually understand site conditions. By freely rotating and zooming a generated 3D model on screen, details that are hard to see in flat photos can be comprehended three-dimensionally. Moreover, by performing cross-sectional cuts on arbitrary planes within the software, longitudinal and cross-sectional views of terrain or structures can be created. Even for complex-shaped members, if a point cloud exists, cross-sectional geometry and thickness can be checked afterward, making comparison and verification with design drawings straightforward.

In terms of data interoperability, photogrammetry-derived 3D data are powerful. Point clouds or models with absolute coordinates align precisely when overlaid on CAD drawings, BIM/CIM models, or GIS maps. For instance, overlaying a point cloud obtained from bridge inspection onto existing bridge ledger drawings allows precise recording of crack locations or member dimensions, and in construction management, design 3D models can be compared with as-built point clouds to detect deviations. The ability to directly incorporate on-site point clouds into design and maintenance digital foundations is a unique advantage of photogrammetry with absolute coordinates.

Recent advances have also addressed data volume. Depending on the acquisition area, 3D models generated from photos often have relatively smaller file sizes compared to raw point clouds from laser scanners. Converting to mesh (polygon) models with texture images can produce files lighter than point clouds, enabling smooth sharing via the cloud and viewing on mobile devices. For these reasons, point cloud data obtained by photogrammetry serve not only as surveying deliverables but also as valuable site records and communication tools.

Photogrammetry Use in Infrastructure

Photogrammetry is expanding its role in infrastructure such as roads, bridges, tunnels, and dams. Traditionally, capturing terrain and structural geometry required significant time and effort, but using photogrammetric point clouds can improve efficiency. For example, in wide-area road planning, photos taken from the air by drone can be processed with photogrammetry to create detailed terrain models for route selection and earthwork planning. Even in mountainous road slopes, high-density point clouds generated from photos enable careful examination of slope gradients and shapes, facilitating extraction of hazard areas and planning reinforcement work.

Photogrammetry is also powerful for detailed recording and analysis of large structures like bridges and tunnels. Creating 3D models of piers and girders enables digital measurement of clearances and displacements. Capturing tunnel interiors and converting them to point clouds allows continuous acquisition of internal cross-sectional shapes, which helps detect cross-section shrinkage or deterioration of lining concrete. When combined with RTK, these models are linked to public coordinates and can be used directly as surveying results, easing comparison with design drawings and existing ledgers.

In infrastructure maintenance, periodic monitoring is essential, and photogrammetry makes it easier to grasp changes over time. For example, in dam displacement measurement, annual point clouds of the dam surface obtained via photogrammetry can be compared through differencing analysis to detect small deformations. Whereas traditional measurements provided discrete data from scattered instruments, point cloud models capture changes across entire surfaces, making them useful for assessing the health of infrastructure.

Photogrammetry Applications in Construction Management

Photogrammetry is a powerful tool for construction management as well. Regular photogrammetric surveys and 3D model generation help manage as-built conditions and visualize progress. For earthworks, conducting drone surveys before and after excavation or embankment and calculating volumes from the respective terrain point clouds enables accurate earthwork management. Tasks that used to involve manual point measurements and sectional drawings can now yield quantities over large areas quickly using point cloud data.

In building and civil engineering works, photogrammetry serves as an excellent record of site conditions. Recording rebar arrangements and formwork installation in 3D models allows later verification of compliance with drawings, and comparing shapes before and after concrete placement helps check construction errors. With RTK-enabled photogrammetry, direct measurement of dimensions or slopes on the model is possible, enabling site supervisors and inspectors to perform quality checks on the spot.

Photogrammetry is also effective for creating as-built drawings after completion. Traditionally, creating as-built (completion) drawings required reconciling design drawings with surveyed as-built measurements, but using point clouds from photogrammetry, the actual 3D model of the finished structure can be preserved as deliverables. This streamlines post-construction documentation and ensures more accurate records for future reference.

Use in Structural Inspection and Maintenance

Photogrammetry offers new approaches for structural inspections of tunnels, bridges, and buildings. Inspection used to rely on visual checks and sounding tests, but digitizing the entire structure with high-precision photogrammetry allows detailed assessment from the office. For example, bridge inspections can generate 3D models of the underside of girders from high-magnification photos to digitally measure and record crack locations and lengths. For tunnel wall crack surveys, continuous imaging while driving through the tunnel can be converted into point clouds and used to color-code deterioration areas—an advanced approach already in practice.

The strengths of photogrammetry-based inspections are improved safety and efficiency. Compared to traditional methods requiring scaffolding or elevated work platforms for high or confined areas, remote digital inspection eliminates the need for personnel to enter hazardous locations. Because the entire structure can be recorded at once, oversights are reduced and variability among inspectors’ evaluations is minimized. Stored data can be compared with future inspections to ensure new deterioration is not missed, enabling quantitative maintenance management.

Of course, photogrammetry inspections are limited to areas accessible to cameras, so internal or hidden deterioration may still require direct inspection. However, combining drones or robots to photograph inaccessible areas and linking inspection findings to the photogrammetric model for visualization expands applications. As smartphone RTK-based easy point cloud acquisition spreads, even small-to-medium-size structures can be regularly archived with high-precision digital records, supporting infrastructure longevity initiatives.

Changing the Field: The Potential of Smartphone RTK Photogrammetry — LRTK

As described above, the fusion of photogrammetry and RTK, and the emergence of smartphone RTK, are making high-precision 3D surveying more accessible than ever. A prime example is the smartphone-based positioning and measurement solution “LRTK.” LRTK attaches an ultra-compact RTK-GNSS receiver to a smartphone or tablet and uses a dedicated app to perform photogrammetry and point cloud scanning, enabling anyone to achieve centimeter-level accuracy (half-inch accuracy) surveying easily.

With LRTK, you no longer need to carry heavy surveying equipment; you can measure coordinates of necessary points or simply walk around a target and take photos with a smartphone to generate high-precision 3D models on the spot. Because the acquired models already have global positioning coordinates, they can be used immediately on drawings or GIS without later alignment on a PC. LRTK also supports offline measurement in areas without internet access because processing can be completed within the smartphone.

Viewing a point cloud model displayed in real time on the smartphone feels like wielding a virtual 3D scanner on site. The generated 3D model can be shared with stakeholders via the cloud so the entire project team has the same understanding of current conditions, and surveying results can be compiled into reports on the same day. The time from surveying to drawing that once took days can be drastically shortened by LRTK.

By lowering the barriers of expensive equipment and specialist knowledge, photogrammetry combined with smartphone RTK is turning 3D surveying into a routine workplace tool. Adopting solutions like LRTK means that in infrastructure inspection and construction management, anyone can immediately acquire high-precision 3D data whenever needed. The efficiency and labor savings that photogrammetry and smartphone RTK bring to the field are immense, and these tools will become indispensable for more and more practitioners. Why not take this opportunity to try cutting-edge smartphone RTK photogrammetry and experience centimeter-level accuracy (half-inch accuracy) 3D surveying for yourself?