Point cloud generation automated too! Promoting DX (Digital Transformation) with drone surveying that doesn't require high-spec PCs

Table of contents

• What is drone surveying

• What is point cloud data

• Why high-spec PCs were required for point cloud generation

• Latest technologies that make high-spec PCs unnecessary

• Benefits of drone surveying without high-spec PCs

• DX promotion effects brought by drone surveying

• Easy drone surveying with LRTK

• FAQ

Drone surveying, which uses drones (unmanned aerial vehicles), has recently attracted significant attention in the surveying and construction industries. By photographing sites from the air, wide-area terrain data can be acquired in a short time, and in terms of safety and efficiency it greatly surpasses traditional surveying methods. However, one of the challenges that hindered the spread of drone surveying was that processing the acquired data required high-performance computers (so-called high-spec PCs). For example, generating detailed 3D models or point cloud data from the many photos taken by a drone demanded PCs with large memory capacity and high-performance GPUs, which could be difficult for small and medium-sized companies or field-level operations to adopt.

Recently, however, new technologies and services have emerged that automate the process of point cloud generation itself and complete it in the cloud. By offloading heavy processing to the cloud, environments are being established where anyone can easily enjoy the benefits of drone surveying without high-spec PCs. In this article, we first clarify what drone surveying is, its advantages, and the traditional challenges, then introduce the mechanisms by which the latest cloud technologies enable point cloud generation without high-performance PCs. We also explain how these technologies contribute to DX (Digital Transformation) in industries such as construction. Finally, using the cloud service LRTK as an example, we introduce concrete solutions useful on-site for easily starting drone surveying. Keep up with the latest trends in drone surveying to help promote DX in your operations.

What is drone surveying

Drone surveying is a new surveying method that equips small unmanned aerial vehicles (drones) with cameras and sensors to photograph and measure the ground and structures from the air. By replacing surveying tasks that people used to perform on the ground with drones, wide-area current-condition data can be acquired quickly and safely. From the large number of photos captured by a drone, detailed three-dimensional terrain models and orthophotos (continuous aerial images viewed from directly above) can be generated, so drone surveying is being applied to a wide range of uses such as terrain surveying at civil engineering and construction sites, management of earthwork volumes, forest surveys, and post-disaster damage assessment.

A major advantage is that drones can easily acquire data from above even in situations that were previously difficult, such as surveying dangerous cliffs or steep slopes that are hard for people to access, or assessing disaster sites. Also, a single flight can acquire a very large amount of information, enabling the creation of high-density point cloud data that captures minute surface irregularities, which in turn allows advanced analyses such as earthwork volume calculations and comparisons with design drawings. Because of this efficiency and high accuracy, drone surveying is expected to become the next-generation standard surveying method.

What is point cloud data

One of the representative deliverables obtained from drone surveying is point cloud data. Point cloud data is three-dimensional data that represents the surfaces of terrain and structures as a collection of countless points. Each point has X, Y, and Z coordinate values (such as longitude, latitude, and elevation), and the aggregation of points digitally reproduces the surface shape in detail. Simply put, it’s like a digital model that replaces the actual terrain with a large number of points.

By analyzing this point cloud, you can grasp the current terrain in detail, calculate distances or elevation differences between arbitrary points, and compute area or volume over a specified range, among various other measurements and analyses. How do we obtain point cloud data from drones? A common method is photogrammetry using the numerous overlapping photos taken by the drone. Software detects feature points in multiple photos and matches them across the images, allowing the 3D positions of each point to be computed inversely. In this way, a three-dimensional collection of points—i.e., a point cloud—is generated from photos.

Alternatively, if a drone is equipped with a laser scanner (LiDAR), it can obtain three-dimensional point cloud data directly by emitting laser pulses toward the ground and measuring their reflections while flying. Both photogrammetry and laser measurement can acquire high-density, high-accuracy point cloud data. In drone surveying, these point clouds are used to draw contour lines, create terrain cross-sections, compare terrain before and after construction, and compile surveying maps and reports.

Why high-spec PCs were required for point cloud generation

To put drone-acquired data to practical use, the process of generating point cloud data from aerial photos is indispensable. However, this point cloud generation stage is extremely computationally intensive, and traditionally required high-performance PCs. For example, when reconstructing a point cloud from more than 100 high-resolution photos, it is necessary to calculate tens of millions to hundreds of millions of points. This demands large memory capacity and high CPU/GPU performance, and typical office PCs could take tens of hours to process or might freeze mid-process.

There were also challenges in displaying and editing the generated point clouds. Because point clouds are very large data sets, specialized 3D viewer or CAD software is required, and running these smoothly also required high-spec PCs. Handling large point clouds often led to insufficient storage or graphics processing power, causing sluggish behavior even when simply scrolling. Thus, fully adopting drone surveying required purchasing expensive workstation-class PCs and specialized software and acquiring the skills to use them, creating significant initial cost and learning hurdles.

Latest technologies that make high-spec PCs unnecessary

In recent years, advances in cloud technology have produced new solutions that realize high-spec PCs unnecessary in drone surveying. A prime example is services that automatically execute point cloud generation from photos in the cloud. Users simply upload the image data captured at the site to the cloud via the Internet, and high-performance machines on the server side automatically perform the processing. The entire sequence—from three-dimensional reconstruction and point cloud generation from hundreds of photos, to the composition of orthomosaic images—can be completed with the push of a button. Complex parameter settings or artisanal software operations are not required.

Because all heavy computational processing can be delegated to the cloud, a standard laptop or even a tablet at the site is sufficient. There is no need to bring high-performance equipment to the field, and data processing can be performed easily on the go. The cloud can utilize server environments with large memory and multiple CPU cores, sometimes delivering results faster than local PC processing. If you upload data at night, a high-precision 3D model may be ready on the cloud by morning, allowing efficient use of time.

Point clouds and orthophotos automatically generated by cloud services are stored in the cloud. Users can view and inspect 3D point clouds through a web browser and perform additional analyses or downloads as needed. Because results can be viewed in the browser without installing software on a PC, it is easy to share data from the field with office staff. Recently, drones capable of high-precision positioning with RTK-GNSS have become widespread, and cloud processing is highly compatible with such RTK data, automatically assigning absolute coordinates (global geodetic system coordinate values) to point clouds based on photo position information. In environments where sufficient accuracy can be obtained, high-precision surveying deliverables suitable for construction control can be produced without the numerous ground control points that were previously required. Of course, if you want to further improve accuracy, coordinate alignment corrections using ground control points can be applied in the cloud, supporting precision management.

Benefits of drone surveying without high-spec PCs

Cloud-based drone surveying offers many benefits to users. Because it does not require expensive equipment or advanced specialized skills as before, the barriers to introduction and operation are lowered, accelerating on-site DX. The main advantages are as follows.

• Reduced initial investment: There is no need to purchase high-performance workstations or specialized software; you can start with the PC and drone you already have. Equipment costs can be greatly reduced, making drone surveying more accessible to small and medium-sized businesses.

• Ease and convenience: Because processing is automated in the cloud, operation is possible without specialized image-processing knowledge or skills. The simplicity of uploading data and receiving results is a major attraction for field personnel.

• Fast results and decision-making: You no longer need to tie up a PC for hours waiting on heavy processing, and you can confirm results in a short time. In some cases, you can obtain point clouds and orthophotos on the same day as the shooting, allowing you to start revising construction plans or preparing reports on-site.

• On-site utilization: You can perform data processing and analysis on a field laptop without returning to an office with a high-spec PC. For instance, even a remote temporary office at a construction site can leverage processing capabilities comparable to headquarters via the cloud.

• Easy data sharing: Since deliverables are stored in the cloud, they can be shared instantly with stakeholders over the Internet. Recipients can view data in a browser without special software or high-spec PCs, enabling easy confirmation of 3D data.

• Scalability: Many cloud services provide large-capacity storage, enabling organizational accumulation and management of point cloud data for each project. There is no need to prepare your own servers, and the service provider handles backups and data protection, offering peace of mind.

In this way, cloud-based drone surveying solutions solve traditional problems around cost, time, and effort, contributing to operational efficiency and sophistication.

DX promotion effects brought by drone surveying

The spread of drone surveying and the cloud utilization that removes the need for high-spec PCs greatly contribute to industry-wide DX (Digital Transformation). By enabling end-to-end processing from surveying to design and construction management as digital data, work flows become dramatically more efficient and advanced. Tasks that were previously done with manual field notebooks or two-dimensional drawings can now reproduce site conditions digitally using 3D point cloud data obtained by drones, enabling swift comparisons with design data and verification of as-built conditions. This allows on-site decisions and reporting to occur in a shorter cycle, improving overall work speed and quality.

Moreover, centralizing data management in the cloud smooths information sharing among internal and external stakeholders, enabling near-real-time collaboration across geographically separate bases. This supports telework and remote construction management and contributes to work-style reform. Introducing advanced technologies such as drones and AI on-site can be a trump card for productivity improvement in the construction and surveying industries, which face chronic labor shortages. In Japan, the Ministry of Land, Infrastructure, Transport and Tourism has been promoting initiatives such as i-Construction and ICT construction, and the use of three-dimensional surveying data is increasingly required, especially for public works. Drone surveying combined with cloud point cloud processing is attracting attention from many companies as a new solution that responds to the demands of the DX era.

Easy drone surveying with LRTK

One example of a cloud service that enables drone surveying without relying on high-spec PCs is LRTK. LRTK is a domestically produced, cloud-based point cloud generation and sharing platform that features proprietary algorithms to automatically generate high-precision point cloud data from drone-captured images. Usage is simple: upload the photo data acquired on-site to the LRTK cloud on the web. Image analysis is performed on the cloud, and detailed 3D point clouds consisting of tens of thousands to tens of millions of points are generated. The generated point clouds are assigned accurate position information (absolute coordinates) based on surveying coordinate systems, allowing immediate use for comparison with design drawings and as-built management.

LRTK provides unlimited cloud storage, so you can store large-scale point cloud data reaching hundreds of GB without worrying about your PC's disk capacity. In addition, the cloud point cloud viewer allows smooth display and manipulation of 3D point clouds in a web browser without installing dedicated software. Even if you only have a standard office laptop, as long as you have a browser you can smoothly view point cloud data consisting of tens of millions of points and perform necessary measurements and analyses on the spot. For example, you can extract arbitrary cross-sections from the point cloud and download them as CAD data (DXF format), or automatically calculate volumes of fill and excavation—advanced functions that LRTK can execute with a single click.

Since the deliverables are stored in the cloud, you can issue a sharing link for stakeholders to view online. Recipients can check 3D point clouds and orthophotos simply by accessing the link, without preparing high-spec machines or special software themselves. This eliminates the need to exchange huge data files via email attachments or USB drives and smooths communication.

Additionally, to supplement data in blind spots that drones cannot capture from above, LRTK can be integrated with a handheld measurement device using a smartphone called "LRTK Phone." Terrain under trees or the backside of buildings that drones cannot capture can be scanned on foot with a smartphone and then integrated into the drone point cloud to build comprehensive 3D data without omissions. The flexibility to combine drones with other devices according to site conditions is another major strength.

By leveraging LRTK in this way, point cloud processing and 3D analysis that previously required specialized equipment and high skills become accessible to anyone. Eliminating the need for high-spec PCs improves on-site responsiveness and, as a result, promotes DX across operations. For those who want to start drone surveying or companies that wish to review and streamline their current surveying processes, LRTK can be a powerful supporter.

FAQ

Q. Is a high-spec PC really unnecessary to start drone surveying? A. Yes, a high-performance PC is not required. If you use a cloud-based point cloud processing service, heavy image processing is all executed on servers on the Internet. On-site operation can be done from standard-performance laptops or tablets via a browser, and result data can be checked online. However, an Internet connection is necessary to upload large numbers of photos. Still, removing the worry of processing stopping due to insufficient PC specs lets you focus on your work with peace of mind.

Q. Can measurement accuracy be ensured with cloud services that automatically generate point clouds from photos? A. Even with automatic generation, high surveying accuracy can be assured. The main factors affecting accuracy are the GNSS (positioning system) installed on the drone and, when needed, ground control points. Using an RTK-capable drone combined with cloud processing provides high-precision coordinates for each point, achieving accuracy that meets public surveying standards. Even with non-RTK drones, coordinate correction using control points can be performed in the cloud post-processing. In short, precision management equivalent to or better than traditional manual methods can be completed in the cloud.

Q. What equipment and preparations are needed for drone surveying? A. Basically, you can start with a drone, an Internet connection, and a contract (or software) for a cloud service. Use a drone equipped with a high-resolution camera. For surveying purposes, a drone with RTK functionality is desirable, though not mandatory. Before flight, complete aviation law procedures and safety measures for the survey area and plan flights to ensure sufficient photos are captured. Then fly to obtain photo data and upload it to the cloud. High-spec PCs and complex software settings are unnecessary, making operation relatively simple.

Q. Will uploading image data to the cloud take a lot of time? A. Indeed, uploading many photos can take some time. For example, hundreds of photos can amount to several GB or more, and depending on the network environment, uploads may take anywhere from several tens of minutes to several hours. However, you can continue other work on your PC while uploading, which is an advantage. If you really don't have time on-site, you can upload after returning to the office. Once upload is complete, the cloud will handle automatic processing, so you can take advantage of efficient workflows like uploading overnight and receiving results by morning. Overall, this often increases the time available for productive tasks compared to running long local PC processes.

Q. What can point cloud data obtained by drone surveying be used for? A. Point cloud data is highly versatile and can be used in many ways. In terrain surveying, you can create contour maps from point clouds; at construction sites, you can calculate volumes for excavation and fill to support quantity management. Comparing point clouds before and after construction enables verification of as-built conditions, and overlaying planned lines on actual terrain allows error checks—making point clouds indispensable for construction management. Recently, efforts have advanced to create 3D models from point clouds and import them into CAD or BIM software for design and construction planning. In short, point cloud data acquired by drones serves as foundational data that digitally links surveying, design, and construction processes. By introducing drone surveying, you can also consider such advanced uses.