Secrets of High-Precision 200 m (656.2 ft) LiDAR Point Cloud Data: Technologies That Support Accuracy and Quality

Table of Contents

• Basics of LiDAR and point cloud data

• Advantages and challenges of 200 m (656.2 ft) long-range LiDAR

• Sensor technologies that support high accuracy

• Techniques to improve point cloud data quality

• Simple surveying with LRTK

• FAQ

Introduction

LiDAR (light detection and ranging) using lasers has become an indispensable technology for handling three-dimensional spatial information. The countless points acquired by LiDAR, known as point cloud data, serve as a means to digitize the real world with high precision and are being applied across a wide range of fields—from civil surveying and construction to urban planning, autonomous driving, and even the creation of metaverses and digital twins.

In recent years there has been growing demand for point cloud data that are both highly accurate and cover wide areas, and LiDAR sensor performance has improved dramatically. This trend is embodied in the keyword "200 m (656.2 ft) LiDAR point cloud data." It refers to high-resolution point cloud data produced by long-range LiDAR capable of detecting targets at distances of up to approximately 200 m (656.2 ft). Measurements of structures or terrain several hundred meters away—previously difficult—are increasingly possible with high accuracy.

However, the mere ability to measure with lasers at long distances does not automatically guarantee accuracy or quality. Obtaining long-range, high-precision point cloud data requires various innovations and "secrets," spanning sensor hardware technologies to data-processing software. In this article, we will explain in an easy-to-understand way the technical elements that support high-precision 200 m (656.2 ft)-class LiDAR point cloud data, including their principles and key points.

Basics of LiDAR and point cloud data

First, LiDAR stands for "Light Detection and Ranging" and is a remote sensing technology that measures distances to objects by emitting laser pulses. A LiDAR sensor emits laser light at high speed and measures the time until the light reflects off an object and returns. By calculating distance from the round-trip time of light, the surrounding environment can be recorded three-dimensionally as a collection of points (a point cloud). Each of these points is a data point that indicates a specific position in space (X, Y, Z coordinates).

A feature of LiDAR measurement is that it can acquire a large number of points instantaneously in millisecond time scales. Because laser light travels in a straight line, it can directly measure distant objects and fine shapes, and the resulting point cloud data have very high spatial resolution. In addition, the reflectance (intensity) of the laser is recorded as associated information for each point, which can be used to infer differences in material or color and to assist in noise removal. Some LiDAR systems are also integrated with cameras to add color (RGB) information to the point cloud, generating 3D data that more closely resemble reality.

The acquired point cloud data can be used not only to precisely reproduce site geometry, but also directly as digital assets for creating survey maps and 3D models, as well as for construction management and navigation of autonomous mobile robots. High-precision LiDAR point cloud data provide detailed and comprehensive information that manual measurements could not offer, greatly contributing to efficiency and sophistication in operations.

Advantages and challenges of 200 m (656.2 ft) long-range LiDAR

One performance metric for LiDAR is "measurable distance," that is, how far away it can measure. With typical laser wavelengths and output powers, the distance that can be safely measured while protecting human eyes is limited, and many conventional commercial LiDAR systems had effective ranges on the order of about 100 m (328.1 ft). In contrast, recently introduced long-range LiDAR sensors can, under certain conditions, measure out to around 200 m (656.2 ft), allowing distant terrain and obstacles to be captured as point cloud data. The arrival of the 200 m (656.2 ft)-class LiDAR has dramatically increased the area that can be measured in a single pass, offering the advantage of efficiently scanning large sites and tall structures.

Situations where long-range LiDAR is useful include: automotive sensors for highways, which require detection of obstacles hundreds of meters away; landslide surveys in mountainous areas and monitoring of dams and bridges, where wide-area measurements from remote points are beneficial; and urban 3D mapping and disaster prevention, where long range helps when surveying from high vantage points to capture an entire cityscape.

On the other hand, realizing a 200 m (656.2 ft) measurement range presents several technical challenges. First, laser light attenuates with distance, and the signal returning to the sensor becomes weak. Ambient light noise (such as sunlight) adds further interference, so detecting faint distant reflections accurately requires high-sensitivity photodetectors and sophisticated signal processing. Also, as measurement distance increases, distance error per point (ranging accuracy) tends to grow. To suppress this error growth, precise timing circuits (clocks) and calibration techniques that are less affected by temperature changes are required.

Moreover, the farther the distance, the larger the spacing between projected points for the same angular resolution (laser scan step), so point cloud density decreases with distance. In other words, scanning an object 200 m (656.2 ft) away may result in coarser point spacing and reduced detail compared to close-range scanning. To address this, the latest LiDAR systems increase the laser pulse repetition frequency to emit more pulses per second, or adopt multi-line configurations that simultaneously emit many beams (channels) — 16-line, 32-line, 64-line, 128-line, etc. — to ensure sufficient point cloud density even at long ranges.

Another major challenge for long-range LiDAR is eye safety. Because lasers are powerful light sources, increasing output too much can pose safety risks to humans. One solution is to switch the laser wavelength to bands that are less likely to reach the human retina (for example, the 1550 nm band in the infrared), allowing higher output light to reach longer distances while maintaining safety. However, changing wavelength also brings challenges such as the need to develop compatible photodetectors and increased costs.

Thus, operating 200 m (656.2 ft)-class long-range LiDAR with high accuracy and quality involves addressing problems through both the evolution of optical and electronic components and clever system design.

Sensor technologies that support high accuracy

To obtain high-precision point cloud data, the LiDAR sensor hardware itself is critically important. One key area is technologies that improve the accuracy of distance measurement. LiDAR calculates distance by measuring the round-trip time of laser pulses, and the timing measurement requires precision on the order of less than a nanosecond (one billionth of a second). To achieve this, high-frequency clocks are used in timing circuits and fast electronic circuits that can amplify and detect tiny time differences are employed. As a result, the latest LiDAR sensors achieve distance accuracy of less than a few centimeters (typically around ±2–3 cm (±0.8–1.2 in)). For example, careful calibration and compensation are applied so that errors remain within the same few centimeters whether at 1 m (3.3 ft) or 100 m (328.1 ft).

Advances in photodetectors and signal processing are also key to improving accuracy. To capture weak reflections from distant targets, LiDAR uses high-sensitivity sensors such as avalanche photodiodes (APD) and silicon photomultipliers (SiPM). These devices internally amplify incident photons avalanche-style, enabling detection of tiny signals. Additionally, in waveform LiDAR—where the entire returned signal waveform is recorded and analyzed—techniques that estimate distance from waveform shape rather than simple peak detection allow sub-pixel-level distance estimation, contributing to improved distance resolution.

There are also mechanical and structural design efforts for high accuracy in sensor mechanisms. Many LiDARs scan the surroundings by steering laser light with rotating mirrors or prisms, and these drive systems require high-precision control to minimize vibration and wobble. Mechanical scanners use high-quality motors and encoders to rotate at stable speeds and suppress angular errors. On the other hand, solid-state LiDARs, which use semiconductor-based, non-mechanical scanning, employ MEMS mirrors or optical phased arrays to steer beams. These have fewer moving parts and excellent durability but require sophisticated implementation to ensure wide field of view and high resolution while maintaining stable operation.

Calibration of the sensor must not be forgotten. LiDAR manufacturers perform thorough calibration during production so that sensor outputs correspond to accurate distances. Temperature testing, compensation for long-term drift, and adjustment of alignment between multiple beams are used to remove hardware-originated errors to the greatest extent, enabling stable accuracy in field measurements. Part of the "secret" of high-precision LiDAR lies in this invisible, thorough technical tuning.

Techniques to improve point cloud data quality

Even when hardware is high-performance, raw point cloud data can contain various noise and errors. Therefore, software techniques to further improve the quality of acquired point clouds are indispensable. One representative technique is noise filtering. LiDAR can detect airborne particles, raindrops, or reflections from glass surfaces, which become unwanted points (noise) in the data. Dedicated algorithms automatically detect and remove points that are abnormally strong or isolated from their surroundings. Because faint noise can more easily be mixed into measured points at long distances, such filtering is important to produce clean point clouds.

Utilizing multiple returns (multi-echo) also helps improve quality. When a single laser pulse partially reflects from multiple objects, the LiDAR sensor can detect not only the nearer reflection but also reflections from objects behind it (first echo, second echo, etc.). For example, when scanning a forest, reflections from leaves as well as from the ground beneath the foliage can be obtained, improving ground surface detection accuracy. Some of the latest LiDAR units can capture more than 10 echoes from a single pulse, generating point clouds that are less likely to miss features even in complex environments.

Combining multiple scan datasets to densify point clouds is also common. For ground-based LiDAR, scanning the same object from different positions and registering them together can produce a high-density point cloud that compensates for occlusions. To ensure registration accuracy, artificial target markers can be placed on site as references, or SLAM (simultaneous localization and mapping) algorithms can be used to incrementally align point clouds. In integrating multiple datasets, high-precision IMU (inertial measurement units) and GNSS position and attitude records can also be useful.

Further, analysis and post-processing of the acquired point cloud contribute to quality enhancement. Beyond removing unwanted points, smoothing random noise (fluctuations) and classifying point clouds by category—such as ground surface or structures—are performed. Classification makes it easier to distinguish noise from real objects, allowing specific classes of points (for example, vehicles or vegetation) to be removed when needed. When converting point clouds to meshes for 3D models, outlier rejection based on distances to neighboring points is incorporated to improve the quality of the final deliverable.

In this way, maximizing LiDAR point cloud data quality requires multilayered adjustments and optimizations from sensor acquisition to data post-processing. High-precision, high-quality point clouds are never simply "as-acquired" but are supported by many technologies that work behind the scenes.

Simple surveying with LRTK

Until now, using high-precision LiDAR point cloud data often required specialized surveyors and advanced equipment, and obtaining and processing data typically took considerable effort and time. Recently, however, solutions have emerged that make these tasks easier to perform. One such approach is "simple surveying with LRTK."

LRTK is a compact, high-precision GNSS receiver developed by Refixia, and it realizes centimeter-level positioning using real-time kinematic (RTK) techniques. By combining this LRTK with LiDAR, on-site surveying work can be greatly streamlined. Specifically, by linking the high-precision self-position information (latitude, longitude, and height) measured by LRTK to the point cloud acquired by the LiDAR sensor, absolute coordinates can be assigned to the point cloud data. This makes it possible to align the acquired point cloud directly to a map coordinate system without complex post-processing.

For example, traditionally, registering point clouds from ground-based LiDAR scans required establishing known control points at each measurement location or spending time later to adjust point clouds. With LRTK, the sensor’s position can be recorded with high accuracy in real time during measurement, allowing geographic reference information to be attached to the point cloud as it is captured. In drone-mounted LiDAR terrain surveys or mobile mapping where vehicles carry LiDAR and LRTK, dynamic measurements—"measuring while moving"—become possible, enabling the creation of accurate 3D maps without placing survey control points each time.

The advantage of simple surveying with LRTK is that it is easy for anyone to handle without specialized operation. By combining a small, portable LRTK receiver with LiDAR equipment, and linking them to dedicated apps or cloud services, point cloud acquisition and positioning can be performed automatically on-site with the press of a button, and the data are immediately visualized as 3D models. Tasks that once depended on veteran surveyors are becoming more accessible thanks to LRTK. To allow more sites to easily enjoy the benefits of high-precision 200 m (656.2 ft) LiDAR point cloud data, LRTK plays a role in lowering the barrier to surveying.

FAQ

Q. What is LiDAR? A. LiDAR (light detection and ranging) is a remote sensing technology that uses laser light to measure distances to objects. The abbreviation stands for Light Detection and Ranging. It emits laser pulses at high speed and calculates distance from the round-trip time of the reflected light. Because it can acquire distance data for many points, it records the surrounding environment as a three-dimensional "point cloud."

Q. How accurate is LiDAR point cloud data? A. It depends on the sensor performance and conditions, but modern high-performance LiDAR often maintains distance errors within about ±2–3 cm (±0.8–1.2 in). High-precision models can measure the distance of each point very precisely and are calibrated so that accuracy is almost uniform from short to long ranges. However, measurement accuracy is influenced by environmental factors (temperature, weather, material of the target, etc.), so these values represent ideal-condition performance.

Q. What are the advantages of LiDAR that can measure up to 200 m (656.2 ft)? A. The main advantage of using long-range LiDAR is the ability to acquire data over a wide area at once. For example, terrain that previously required multiple relocations to measure can be covered from fewer setup locations with a 200 m (656.2 ft) range LiDAR. In autonomous vehicles, long-range detection contributes to safety by enabling earlier detection of distant obstacles. Long-range LiDAR is also useful for measuring tall buildings, mountainous areas, and for grasping wide-area situations during disasters.

Q. Do weather conditions such as rain or fog affect LiDAR measurements? A. Yes, they do. Raindrops and fog droplets scatter and absorb laser light, which can increase noise in point cloud data and reduce effective range. Under heavy rain or dense fog, acquiring distant point clouds may be difficult. However, some of the latest LiDARs have features to reduce weather noise or use wavelengths and output settings that operate under certain adverse conditions. It is desirable to choose measurement times with the best possible weather conditions.

Q. What is simple surveying with LRTK? A. LRTK is a high-precision RTK-GNSS receiver, and the method refers to combining it with LiDAR to perform easy surveying. Traditionally, aligning point cloud data to map coordinates required placing control points or post-processing, but by using LRTK you can obtain the surveyor’s high-precision position in real time and assign absolute coordinates simultaneously with point cloud acquisition. In this sense, using an LRTK-enabled system allows easy, high-precision 3D surveying on site even without specialized technicians.