How to perform 3D measurement without markers? 6 steps to avoid failure on site

In recent years, opportunities to use 3D measurement technology for as-built management and inspection tasks at construction and civil engineering sites have been increasing. Efforts to acquire entire sites as point cloud data and use them for construction management are spreading through drones, laser scanners, and photogrammetry.

When performing such 3D measurements, it used to be common practice to place markers (control points) on site to ensure measurement accuracy. However, installing and measuring markers requires effort and time, and depending on the terrain and environment, it may be difficult to place them as intended.

Recently, methods that perform 3D measurement without markers have attracted attention. Advances in technology have made it increasingly possible to acquire high-accuracy point cloud data without placing survey targets on site.

This article explains what “markerless” 3D measurement is and introduces six steps to avoid failure on site. Considering various use cases such as drone surveying, terrestrial laser scanning, photogrammetry, and mobile mapping, we will look in detail at points to keep site work running smoothly.

Table of contents

• What markerless 3D measurement is

• Differences from traditional methods that use markers

• 3D measurement technologies that enable markerless operation

• Step 1: Preplanning and equipment preparation

• Step 2: On-site environment check and device setup

• Step 3: Data acquisition (photography/scan) points

• Step 4: On-site checks after data acquisition

• Step 5: Data organization and backup

• Step 6: Processing and accuracy verification

• The future of markerless 3D measurement and use of LRTK

What markerless 3D measurement is

“Markerless 3D measurement” means, as the name implies, conducting three-dimensional surveying and measurement without installing reference markers (targets or control points) on site. Normally, in photogrammetry you place known ground control points (GCPs) on the ground to serve as references during image analysis, and in laser scanner measurement you place reflective targets in multiple locations to align each scan. In markerless measurement, you obtain 3D data and determine model dimensions and coordinates without using such artificial markers at all.

Instead of markers, advanced hardware and software processing realize high accuracy. For example, if accurate position information from RTK-GNSS is attached to each measurement dataset, photos and point clouds can automatically be given geolocation coordinates so control points are unnecessary. Also, when combining multiple scan datasets, targetless registration (automatic alignment without targets) can align point clouds based on matching overlapping feature points, integrating them into a single model without relying on markers. Furthermore, in mobile mapping where measurements are taken while moving, SLAM (simultaneous localization and mapping) algorithms estimate the device’s position in real time while acquiring point clouds, eliminating the need for pre-established control points. In this way, a combination of high-precision positioning hardware and software processing enables markerless 3D measurement.

Differences from traditional methods that use markers

Traditional 3D measurement methods that use markers required placing control points or targets in advance and measuring their coordinates with instruments like total stations. In photogrammetry, control points had to be arranged so that multiple GCPs appeared in aerial images, and in point cloud registration a common target (prism or reflective sticker) had to be visible in each scan to link the datasets in post-processing. These tasks required going around the site to place markers over wide areas, consuming manpower and time. On large sites or in areas with poor access, it could be difficult to position control points properly, causing uneven measurement coverage and raising safety concerns. Moreover, even if high-quality 3D data were obtained, surveying errors on control points or forgetting to place targets could cause positional shifts or scale mismatches that might render the measurement invalid.

On the other hand, markerless methods eliminate such cumbersome preparatory work and greatly improve on-site efficiency. You can start measurements quickly even on large sites and avoid sending personnel into hazardous areas to place markers. For example, with an RTK-enabled drone survey, simply launching the drone and flying an automated route can attach high-accuracy positioning to aerial photos, eliminating the need to install ground control points that were previously essential. Terrestrial laser scanners can also register point clouds without targets as described below, removing the need to apply reflective stickers at multiple locations from the outset. Thus, markerless measurement significantly reduces the manpower and time required for marker placement while still delivering high-accuracy survey results.

However, because markerless methods do not rely on markers, the performance of the equipment and the on-site measurement procedures have a greater impact on accuracy. If you begin shooting before GNSS positioning is sufficiently accurate, or attempt to register point clouds with insufficient overlap, the resulting model may have misalignments. The accuracy management that markers traditionally provided must instead be supplemented by equipment setting checks and data quality inspections. In other words, appropriate procedures and check systems on site become even more important for markerless 3D measurement. The following explains the specific steps in six points.

3D measurement technologies that enable markerless operation

Key technologies that make markerless 3D measurement possible include the following:

Drone surveying: This method uses UAVs (drones) to take photographs or laser scans from the air. RTK-GNSS-equipped survey drones have become widespread, recording centimeter-class position information for each aerial image. Therefore, without installing ground control points, you can generate high-accuracy 3D models (orthomosaics and point clouds) from acquired images. Even for large-area terrain surveys, data can be acquired quickly as long as takeoff and landing space is ensured, greatly reducing the need for on-ground marker placement.

Terrestrial laser scanner: Devices mounted on tripods perform 360-degree laser measurement from the ground. Typically, when scanning from multiple positions, reflective targets visible from each position are placed and used to join point cloud datasets. However, modern scanner workflows can automatically align scans in post-processing without targets if there are clearly overlapping surface geometries. If adjacent scans have sufficient overlap, you can combine point clouds with high accuracy without the hassle of placing targets. Depending on site conditions, you can also apply known coordinates to some characteristic points afterwards to translate and rotate the whole model, but the measurement flow itself can be completed without marker placement.

Photogrammetry (Structure from Motion): This method uses digital cameras or smartphones to photograph an object from multiple directions and creates 3D models using SfM (Structure from Motion). Traditionally, scale bars or known points needed to be placed on site to give the final model an accurate scale and position. Nowadays, however, built-in smartphone sensors or external GNSS can record camera positions, allowing absolute coordinates to be assigned to models. For example, using an RTK-GNSS device attached to an iPhone can record camera positions for each photo with an accuracy of several centimeters (cm level accuracy, half-inch accuracy). Using that data, you can generate full-scale point cloud models aligned to a geodetic coordinate system without control points, greatly expanding the uses of site photographs.

Mobile mapping: Mobile mapping technologies measure while a person walks or a vehicle moves. Vehicle-mounted MMS (Mobile Mapping Systems) combine high-precision GNSS and IMU-equipped laser scanners to turn the surroundings into point clouds simply by driving. For roadways and long tunnels, this eliminates the need to close the road and place markers and significantly shortens work time. Handheld and backpack-mounted mobile scanners have also appeared; SLAM enables indoor and outdoor 3D scanning simply by walking around. Because these systems estimate their position while generating maps in real time, they can acquire continuous point cloud data without pre-placed reference markers.

Step 1: Preplanning and equipment preparation

Successful 3D measurement requires planning before going to the site. First, clarify the measurement objectives and required accuracy. The appropriate measurement method and equipment differ depending on whether you need terrain-wide surveying or detailed shape recording of a structure.

For example, choose drones for wide-area terrain surveys, terrestrial laser scanners for high-precision local measurements, and smartphone photogrammetry for convenience—select the method appropriate to the site conditions and required accuracy. Also survey the measurement area in advance and plan drone routes and scanner placement. For aerial photography, consider flight altitude and overlap; for laser scanning, think about scanner positions and the number of setups; for photogrammetry, determine shooting point spacing—plan thoroughly so that no blind spots or data shortages occur later.

Once planning is solidified, carefully prepare the equipment you will use. For drones, charge batteries, prepare spares, and confirm transmitter and RTK base station settings. Calibrate and test laser scanners and GNSS devices beforehand to avoid errors on site. If using a smartphone for photogrammetry, ensure adequate storage and test the app beforehand. Also remember to obtain necessary permissions and adjust schedules as required. Clearing regulatory requirements such as drone flight permits or night-work notifications before the day of work reduces the risk of a “failed plan” on site. Identifying potential issues during preparation and securing the necessary equipment, personnel, and time leads to smoother measurement on the day.

Step 2: On-site environment check and device setup

Upon arrival, first check the surrounding environment and prepare conditions so measurements can be taken safely and appropriately. If there are general vehicles or workers within the measurement area, temporarily restrict access or ask them to stay away from the target area. Moving objects such as people or heavy machinery appearing during photography or scanning introduce blur and noise into the data and reduce processing accuracy. Also remove unnecessary obstacles (equipment or temporary structures covering the target) so the whole target is visible to cameras and lasers. Choose a stable location so tripods do not tip over on poor footing, and implement fall-prevention measures on slopes or at heights.

Next, perform initial settings and operation checks for the equipment on site. If using RTK-GNSS, set up the base station in a location with an open sky view and establish accurate reference coordinates. For network RTK, connect smartphones or drones to correction services such as Ntrip and confirm that augmentation signals are being received. GNSS can take several minutes to acquire satellites, so ensure you have a Fix solution (centimeter-level positioning) before starting shooting or scanning. For drones, perform compass calibration and gimbal operation checks before takeoff and confirm that the home point (takeoff location) is recorded correctly. For laser scanners, verify leveling with a level and, if necessary, set a reference direction (when electronic compasses or backsight functions are available). For cameras, set focus and exposure appropriately and take test shots to ensure images are sharp. Thoroughly performing these initial checks prevents device malfunctions and data deficiencies during measurement.

Step 3: Data acquisition (photography/scan) points

Now begin acquiring the 3D data. For photogrammetry, photograph the subject and the entire site from various angles according to plan. For drones, fly at a stable altitude and speed along the planned route and secure sufficient overlap (aim for 70% or more both front-to-back and side-to-side). To prevent blur and out-of-focus shots, use a faster shutter speed and increase ISO as needed. When shooting handheld from the ground, move around the subject to shoot from all sides and capture both near and far views to improve accuracy. Pay special attention to recessed parts of structures and areas prone to shadows—shoot from multiple angles so gaps do not appear in processing. Consider weather and sunlight; if strong shadows from direct sunlight are problematic, change angles or shoot under cloudy conditions or at different times of day.

For laser scanning, execute scans sequentially at the predetermined measurement points. Set each scan to capture as wide an area as possible so adjacent scans have adequate overlap. For example, when scanning around a building, place tripods so that the field of view overlaps at least 30% with neighboring scans.

After scanning, preview the point clouds on site and check for obvious noise or missing areas. When using mobile instruments, walk slowly and thoroughly so the path has no gaps, and if necessary retrace steps to pass the same location multiple times. Because SLAM methods can accumulate positional errors gradually depending on the environment, performing a loop closure—returning to the start point—can effectively improve accuracy. For any method, monitor device behavior during measurement for abnormal error messages or unexpected stops, and if a problem occurs calmly reset the device or power-cycle it before resuming data acquisition.

Step 4: On-site checks after data acquisition

When all photography and scanning are complete, perform a quick on-site check of the data. For photogrammetry, import the captured images to a PC or tablet and try aligning a subset of photos in the software if possible. Even if you cannot do full processing, at least display all photos in a list to visually check for blur, exposure errors, or coverage gaps. If you find missing or unclear images, perform supplemental shooting before site conditions change. This step is particularly important for complex targets like bridges or plant facilities where re-acquisition can be difficult.

For laser scanners, if possible, try a simple on-site registration of scans. Some modern models provide tablet apps that allow provisional merging of multiple scans on site and can reveal scan omissions or misalignments. Even without such features, overlaying previews of each scan in sequence helps identify clear gaps or inconsistencies. If you suspect a significant misalignment, perform additional scans from intermediate points or attach targets to characteristic features for re-measurement as a recovery measure. For mobile mapping, review the real-time generated point cloud map along the entire walking route to check for omissions. In all methods, only when you judge the data to be free of problems at this on-site check should you pack up. Being able to address deficiencies on site drastically reduces the risk of returning to the office only to find the data unusable.

Step 5: Data organization and backup

After finishing measurements on site, organize and store the acquired data within the same day. First, transfer all captured images and scan data from devices to a computer and sort them into project folders. If you measured multiple sites consecutively, include site names and dates in folder names to avoid confusion. For photos, create subfolders by shooting location and confirm that sequence numbers are complete to ensure there are no missing or duplicate files. Rename laser scan point cloud files so that station numbers and directions are clear, which will facilitate later registration. If you kept field notes (e.g., “Scan 3 has a lot of noise from high-voltage lines”), save them as text files within the data folder.

After organizing, always create backups. Do not rely solely on SD cards or USB memory devices—copy data to multiple storage destinations such as the PC’s internal drive, external hard drives, and cloud storage. Ensuring that one copy can be restored if another fails is critical. For large aerial photo datasets, watch for transfer errors and confirm that file counts and sizes match the originals. Once backups are complete, data organization is provisionally done. Securely storing the valuable data you worked hard to acquire on site allows you to proceed to the processing stage with confidence.

Step 6: Processing and accuracy verification

Finally, process the acquired data in analysis software to create 3D models and verify accuracy. For photogrammetry, load photos into dedicated SfM software to generate point clouds and mesh models. If RTK-GNSS recorded camera positions during shooting, use that information to assign absolute coordinates to the model. For laser scans, integrate multiple point clouds in software. If you did not perform provisional merging on site, execute automatic registration using feature matching or ICP algorithms to bind all scans into a single coordinate system. At this stage, make minor adjustments in software as needed and check for obvious misalignments. For mobile mapping data, optimize SLAM trajectories with dedicated tools to reduce point cloud distortion.

Accuracy checks of the generated 3D model or point cloud are indispensable. First, measure known distances or heights on the model to see if they are correctly reflected. If you have lengths measured directly on site (for example, building width or road width), measure the distance between the corresponding points in the point cloud to determine the level of agreement. Comparing height differences between arbitrary points with leveling survey values or overlaying model cross-sections with design drawings are also effective. For RTK-GNSS-based measurements, verify that the data aligns with the reference coordinate system (for example, that it matches the public coordinate system). If these checks reveal no issues with accuracy, processing is complete. Using the obtained 3D data to produce required drawings and quantity calculations means your markerless 3D measurement mission has succeeded.

The future of markerless 3D measurement and use of LRTK

Markerless 3D measurement methods are expected to become increasingly widespread. With the trend toward ICT construction and i-Construction, demand for technologies that can digitize sites simply and quickly is rising. As introduced in this article, by following appropriate procedures it has become possible to record site conditions in 3D at practical accuracy levels without placing markers. These new methods, which directly reduce site burden and improve safety, are becoming a standard alternative to traditional surveying.

One example of a tool that makes markerless measurement easy is the ultra-compact RTK-GNSS receiver in the LRTK series, which can be attached to iPhones and iPads. For example, the “LRTK Phone” is an integrated positioning device for smartphones that, when combined with a phone’s built-in camera and LiDAR sensor, can acquire high-accuracy point cloud data without control points. A dedicated app can perform photo measurement and position recording simultaneously, enabling even non-experts to perform site 3D measurement solo. Initial adoption costs are kept low, and use is expanding at small- to medium-scale construction sites. If you feel there are issues in your current surveying workflow, it is well worth trying such modern technologies. Utilize solutions like LRTK to achieve markerless, easy, and high-accuracy 3D measurement, and bring new efficiency and digital transformation to your site operations.