7 Ways to Create Point Cloud Data | Comparing Smartphones, Drones, and Ground-based Surveying

Table of Contents

• 1. Terrestrial Laser Scanner (TLS)

• 2. Laser-scanner-equipped drone

• 3. Photogrammetry (photogrammetry)

• 4. Mobile Mapping System (MMS)

• 5. Handheld 3D scanner

• 6. LiDAR-equipped smartphone

• 7. Acoustic sounding

• Comparison of smartphone, drone, and ground-based measurements

• Summary

In recent years, with advances in laser measurement technology and drones, the use of point cloud data (data that represents objects with countless 3D points) has been spreading at construction, civil engineering, and surveying sites. Point cloud data is digital information that records objects such as buildings or terrain with a large number of measured points (X, Y, Z coordinates) and can reproduce them in 3D space. By using point cloud data, it is possible to safely survey without personnel entering hazardous sites and to greatly shorten work time and improve efficiency. Furthermore, high-precision 3D models and drawings can be created from acquired point clouds, and their use in design and construction management is also progressing. For example, the applications of point cloud data are diverse, including as-built records of civil engineering works, calculation of land volumes, displacement monitoring of infrastructure facilities, digital recording of cultural properties, and surveying forest resources.

On the other hand, there is more than one way to create point cloud data, and the situations in which each method excels, the accuracy, and the costs vary depending on the equipment and techniques used. Some people may worry, "I don't know which method to choose" or "Isn't it difficult without expensive specialized equipment?"

In this article, we introduce seven representative ways of creating point cloud data (acquisition methods) and compare the differences among smartphone, drone, and ground-based surveying. By understanding the characteristics and the advantages and disadvantages of each, please use this as a reference when considering which 3D surveying method is suitable for your company’s site.

1. Terrestrial Laser Scanner (TLS)

Terrestrial laser scanners (TLS) are high-precision laser measuring instruments mounted on tripods and similar supports. They emit a laser beam that rotates 360 degrees, scanning surrounding terrain and structures as surfaces and recording them as countless points. Unlike conventional surveying, which requires measuring point by point over time, a single setup can measure large areas in a short time. Recent TLS units can take millions of measurements per second, allowing fine site conditions to be recorded quickly. Depending on the model, measurement accuracy reaches the millimeter level (mm (in)), making them suitable for acquiring detailed 3D data in places that are difficult to measure from the air, such as tunnel interiors and building interiors.

However, because TLS takes measurements from a fixed installation, blind spots occur where the laser cannot reach from a single position. To capture the entire large site as a point cloud without omissions, the equipment must be moved to multiple positions to scan each area, and the data must be merged (registration) afterward. In addition, the equipment itself is large and expensive, and its operation and data processing require specialized knowledge. The effort involved in setup and relocation also reduces efficiency on vast sites or steep slopes, so it is not necessarily the most suitable method. Still, the high-precision point cloud data obtained with TLS can be used for advanced analyses—such as displacement monitoring of bridges and plant equipment and the creation of BIM models for existing structures—and thus provides value that is difficult to obtain with other methods.

2. Drones Equipped with Laser Scanners

A laser-scanner-equipped drone is a method that mounts a lightweight LiDAR sensor on a small unmanned aerial vehicle (drone) and performs laser measurement of the ground from the air. While flying, it emits lasers toward the ground and structures and measures distance from the time difference of the reflected light, thereby acquiring point cloud data of the terrain. A major advantage is that it can efficiently survey wide areas from the air in a short time, allowing safe assessment of conditions in locations that are difficult for people to enter, such as steep slopes, riverbeds, forests, and disaster sites. In addition, during measurement the drone’s position and attitude are recorded simultaneously with high-precision GNSS and an inertial measurement unit (IMU), and by combining those data the acquired point cloud is assigned accurate Earth coordinates.

On the other hand, in flight the airframe is susceptible to vibration, oscillation, and weather effects, so the accuracy and point density of the data that can be acquired tend to be lower than those of ground-based measurements. In addition, equipment costs are high, and there are operational hurdles such as obtaining flight permissions under aviation law and acquiring drone piloting skills. In situations that require high accuracy, it is important to use each method according to the objective—for example, by combining aerial data with detailed ground surveys as needed. In recent years, drone LiDAR has also been used for post-disaster area mapping and forest resource surveys, and its advantage of being able to acquire detailed terrain models from the air in a short time has been recognized.

3. Photogrammetry (Photogrammetry)

Photogrammetry is a technique that captures many photographs of a subject from various angles with a camera and reconstructs its three-dimensional shape from that image data to convert it into a point cloud. By analyzing multiple photos with dedicated software, the object's three-dimensional form is computed and point cloud data or 3D models are generated. Because it is based on photographs, a major advantage is that it can reproduce not only the object's shape but also surface patterns and colors exactly as they appear in reality. Capturing the ground from the air with a drone can produce wide-area terrain models, and using a camera on the ground or indoors can record building exteriors and room interiors as colorized point clouds.

However, parts not captured in photos cannot be reconstructed, so point clouds cannot be obtained for, for example, the back side of a building, areas in the shadow of an object, or under trees. Also, to obtain high-accuracy point clouds, it is necessary to capture sufficiently high-resolution photos without gaps, and on surfaces with little texture such as featureless walls, or highly reflective or transparent surfaces like glass or water, detection of feature points is difficult and accuracy can become unstable. When measuring deep forest interiors or enclosed spaces, photogrammetry alone is often difficult, so measures such as combining it with laser scanners are required. Furthermore, post-processing of the captured data requires advanced computation, and the more photos there are, the more time and computing power are required. To improve accuracy, it is also common to perform ground control by installing known control points (targets) on site to provide reference coordinates to the photos and point clouds. In addition, because adjacent photos must have 80% or more overlap, it is important to plan the shooting systematically. In recent years, efforts to equip drones with RTK-capable high-precision GNSS to correct capture positions and improve the accuracy of photogrammetry have also become widespread.

4. Mobile Mapping System (MMS)

Mobile Mapping Systems (MMS) are systems that equip vehicles with 3D laser scanners, cameras, high-precision GNSS, and other sensors to measure surrounding terrain and structures while driving. Lasers mounted on the vehicle roof continuously scan roadside and urban structures, allowing efficient acquisition of large-scale point cloud data. Because they can measure wide areas in a planar manner while moving—essentially creating maps as they go—they make it possible to quickly capture road surface geometry, dimensions of tunnels and bridges, and the arrangement of street trees and buildings.

However, introducing MMS requires specialized vehicles and expensive surveying equipment, so the high initial costs and operational hurdles are undeniable. Also, when surveying while moving, integration with high-performance GNSS and gyro sensors is indispensable for precise positioning, and data processing becomes large-scale. Nevertheless, it has attracted attention as a technology that can dramatically improve the efficiency of road and urban surveys that previously required significant manpower and time, and further use is expected in the future. In practice, MMS has begun to be used in creating and maintaining road registers and in constructing 3D maps of urban spaces. By extracting road surface conditions and the placement of signs and guardrails from the acquired point clouds, it can support efficient maintenance and planning. In addition, future applications such as the development of high-precision 3D maps for autonomous vehicles are also anticipated. Note that point cloud data acquired by MMS can reach the scale of several billion points, so the volume of data to be handled can become extremely large.

5. Handheld 3D Scanner

A handheld 3D scanner is a compact 3D measurement device that can be held and used by hand. As the operator walks or moves around the object, it scans the shape in real time by projecting lasers or structured light. Because the device is compact, it is easy to handle even in confined indoor spaces or complex piping installations, allowing for maneuverable measurements. By scanning slowly at close range, it can record detailed point cloud data even for objects with many irregularities or in intricate sites. In addition, many handheld scanners can be paired with tablet devices to review measurement results on the spot and immediately fill in any missing data.

However, handheld scanners are limited to areas within the operator’s reach, so they are unsuitable for wide-area surveying such as entire buildings or large-scale terrain. Also, because they are operated by hand, walking-induced sway and device shake easily affect accuracy, and obtaining stable point clouds requires skill. For high-elevation measurements or long-distance movement, tripod-mounted systems or drones are more efficient, and handheld units should be regarded as devices that deliver their strengths only within a limited range. Many of these handheld scanners utilize SLAM (simultaneous localization and mapping) technology, automatically estimating their position and the surrounding geometry while moving to build point clouds. For example, in situations that are difficult to measure precisely by hand—such as as-built measurements for interior renovation or the digital documentation of cultural properties—these handheld scanners demonstrate their power.

6. LiDAR-equipped smartphones

In recent years, smartphones (mainly the latest iPhone and iPad series) have come standard with a laser ranging sensor called LiDAR, making it easy to perform 3D scans and acquire point cloud data. With a dedicated measurement app, you can simply walk around holding the phone in one hand and capture nearby structures such as walls, floors, and the ground as point clouds in real time. The acquired data can be immediately displayed and saved on the smartphone, and you can measure distances and areas on the spot or share them as 3D models. The biggest appeal of smartphone-based measurement is that anyone can get started without expensive specialized equipment, and it is attracting attention in the context of ICT-enabled construction and the promotion of DX (digital transformation) in the construction industry.

However, there are limits to measurements using LiDAR built into smartphones. Because the range the laser reaches is limited to a few meters (a few ft), it is not suitable for large-scale surveying. Also, compared to dedicated laser scanners, the accuracy and point density of the point clouds obtained are inferior. Furthermore, point cloud data acquired by smartphones do not have absolute coordinates based on a geodetic datum, so when using them as surveying deliverables it is necessary to align them afterward to known control points. At present, it is difficult to use a smartphone alone in situations that require millimeter-level accuracy (0.04 in). Even so, their ease of use and low cost have promoted their use for simple on-site measurements, preliminary inspections, and as-built management support, and they are beginning to demonstrate sufficient practicality for small sites. Moreover, attempts have emerged to combine smartphones with external high-precision GNSS receivers to supplement positioning accuracy, and techniques for achieving surveying-grade results with simple equipment are developing.

7. Acoustic sounding

Acoustic sounding is a method that uses sound waves to measure underwater topography. An acoustic device called a sonar is mounted on a ship or boat and emits sound waves into the water to measure the shape of the terrain. The time between emission and the return of the sound waves after they hit the lakebed or seafloor is measured, and from that time difference water depth and the surface irregularities of the terrain are obtained as point cloud data. Conventional single-beam (single beam) systems have a narrow insonified area and can cover only a limited range at a time, so it was necessary to scan by making repeated passes over the same area. In contrast, modern multibeam sonar, which has become mainstream in recent years, can simultaneously emit many sound waves in a fan-shaped pattern, allowing a wide swath to be surveyed in a single pass and enabling the acquisition of high-density seafloor point clouds in a short time.

However, acoustic waves are easily affected by water properties (salinity, temperature, turbidity, etc.), so accurate data may not be obtainable in deep locations or in environments with poor water quality. Also, multibeam echosounders are expensive and large, so it should be noted that acquisition costs can be very high. Underwater point cloud acquisition is a specialized field, but it is an indispensable technology for water-related civil engineering projects such as river dredging plans, sedimentation surveys of dam lakes, and pre-construction surveys for harbor works. Note that shoreline topography and structures above the water surface cannot be captured by echosounding, so they need to be integrated with land survey data. By combining land-based laser scanning and drone photogrammetry, it is also possible to construct a continuous 3D model from the land to underwater.

Comparing the differences between smartphone, drone, and ground measurements

Methods for acquiring point cloud data can be broadly classified into three types: easy measurements using smartphones, aerial measurements using drones, and measurements conducted by installing or mounting equipment on the ground (ground-based surveying). Because they use different platforms, each has different suitable uses and performance characteristics. In general, in terms of accuracy and point density, ground-based surveying is the highest, followed by drone surveying, and then smartphone surveying. While ground-based equipment can obtain highly detailed data, drones and smartphones offer a trade-off in favor of convenience and speed.

The greatest advantages of smartphone measurements are the ease of use of the device and its low cost. Because they are easy for anyone to handle and require minimal preparation, they are suitable for small-scale surveys, indoor measurements, and preliminary site inspections. On the other hand, the measurable range is narrow and accuracy is limited, and the reliability of the acquired data does not match that of expensive surveying equipment.

Drone surveying can cover large areas in a short time and is excellent for acquiring data in areas that people cannot approach. Whether using photogrammetry or LiDAR-equipped platforms, observing from above is a suitable method for grasping the overall terrain. However, there are operational constraints such as batteries and flight permits, and it is also affected by weather conditions such as wind and rain. In addition, areas that drones cannot access, such as the back sides of structures or indoor spaces, cannot be measured, so it may be necessary to combine drone surveying with other methods as appropriate.

Terrestrial surveying (stationary laser scanners, MMS, handheld devices, etc.) enables precise data acquisition and offers excellent point-cloud accuracy and density. It is an indispensable method when high precision is required, such as recording detailed dimensions inside tunnels or of buildings. However, the area that can be measured at one time is limited, and covering large areas requires manually moving the equipment. Procurement costs and the difficulty of handling the equipment can also make it excessive depending on the scale or purpose of the site.

In terms of upfront costs, smartphone-based measurements are the least expensive, and drone-based photogrammetry can also be started relatively cheaply. By contrast, stationary and drone-mounted laser scanners have high equipment prices, and MMS that requires a dedicated vehicle entails by far the largest cost burden.

Furthermore, each method is also affected by surrounding environmental conditions. In rain, dense fog, or dusty conditions, noise in the laser scanner’s reflection data increases, and photogrammetry is also affected by lighting and weather. Selecting the optimal measurement timing and equipment according to the site conditions will be the key to success.

Note that in actual projects it is common to combine multiple methods to complement each other's shortcomings—for example, using drone photogrammetry to capture wide-area terrain and terrestrial (ground-based) laser scanning to measure important structures in detail. By understanding the characteristics of each technique and combining them according to the site, it becomes possible to acquire 3D data more efficiently and with higher accuracy.

Summary

Above, we reviewed the characteristics and caveats of seven representative methods for acquiring point cloud data. Each method has areas where it excels, so it is important to choose based on a comprehensive assessment of site conditions, required accuracy, cost, and other factors. Using an appropriate method can greatly contribute to improved productivity and enhanced safety in surveying operations.

In recent years, technological innovation has brought forth solutions that make 3D surveying, which previously required specialized equipment, easier to perform. For example, by combining a smartphone’s LiDAR measurements with the iPhone-mounted GNSS high-precision positioning device "LRTK", anyone can obtain centimeter-level (cm) high-precision point clouds (half-inch accuracy) with a smartphone. The ease of smartphone measurements is retained, while real-time position correction provides undistorted, accurate data, allowing users to overcome precision-management challenges that were previously difficult.

Advances in cloud software and AI automated analysis technologies for efficiently processing and utilizing acquired point cloud data have also been remarkable, and the barriers to making use of point cloud data are steadily coming down. If an environment is put in place where your own staff can quickly collect point cloud data on site without having to prepare expensive laser scanners or drones, operational efficiency will improve dramatically. The use of point cloud data is expected to advance further going forward. These 3D measurement technologies are also likely to become an important pillar supporting DX in the construction industry. Why not start with small-scale sites and try these cutting-edge technologies, such as smartphone + LRTK, to help improve site safety and productivity?