3D LiDAR Glossary: Key Terms and Definitions

GLOSSARY TERMS

A

Angular Resolution

Angular resolution is the minimum angle between two points that a LiDAR sensor can distinguish as separate objects. It is typically expressed in degrees. Higher angular resolution produces denser point clouds, improving the ability to detect small objects or distinguish closely spaced targets. For perimeter security applications, fine angular resolution reduces missed detections along fence lines and around obstacles.

---

B

Beam Divergence

Beam divergence describes how much a laser pulse spreads as it travels away from the sensor. A tightly focused beam with low divergence maintains spot size at long range, preserving point density and detection accuracy at distance. High divergence causes the laser spot to grow, reducing energy density and detection reliability at extended ranges. Quanergy Q-TRACK sensors are designed to maintain effective beam characteristics at the ranges required for outdoor perimeter and infrastructure monitoring.

Beam Steering

Beam steering is the method a LiDAR sensor uses to point its laser across a scene. It is one of three independent design choices in any LiDAR, alongside how distance is measured and how returning photons are detected. Common beam steering approaches are mechanical scanning, MEMS micro-mirrors, and optical phased arrays. Flash LiDAR is the exception, illuminating the whole scene at once without steering a beam. Because these three choices are independent, a sensor can combine any steering method with any detector and any ranging method. This is why terms such as MEMS, SPAD, and solid-state are not alternatives to one another: they describe different subsystems.

Bounding Box

A bounding box is a 3D rectangular volume drawn around a detected object in a point cloud to represent its spatial extent. In LiDAR-based tracking, each tracked object is assigned a bounding box with dimensions (height, width, depth) and a position in 3D space. Bounding boxes enable the system to measure object size, estimate type (person vs. vehicle), and pass structured data to downstream analytics or alert systems.

---

D

Detection Range

Detection range is the maximum distance at which a LiDAR sensor can reliably detect and return a signal from a target object under specified reflectivity and environmental conditions. Manufacturers typically specify range against a surface with a defined reflectivity percentage (commonly 10% or 80%) in clear atmospheric conditions. For security and perimeter applications, effective detection range determines how many sensors are needed to cover a given site perimeter.

Direct Time-of-Flight (dToF)

Direct time-of-flight measures distance by timing how long an emitted laser pulse takes to travel to a surface and return. Because light travels at a known constant speed, the round-trip time yields distance directly. dToF is the standard approach in long-range 3D LiDAR, because a short high-energy pulse can be detected reliably at distance. Quanergy Q-TRACK sensors use direct time-of-flight ranging.

---

E

Edge Processing

Edge processing refers to running object detection, classification, and tracking algorithms directly on or near the LiDAR sensor rather than transmitting raw point cloud data to a central server. This reduces network bandwidth requirements, cuts latency to near real time, and allows the system to generate structured alerts without cloud dependency. Quanergy Q-SHIELD supports on-sensor zone configuration, enabling detection zones and alert rules to be defined and executed directly on the sensor without dependence on a central processing server.

---

F

Field of View (FOV)

Field of view is the angular range a LiDAR sensor can scan, expressed as horizontal FOV and vertical FOV. A 360-degree horizontal FOV sensor covers a full circle around the sensor, while a narrower FOV sensor covers a specific sector. Vertical FOV determines how many laser channels are distributed across the elevation angle and affects coverage of ground-level and elevated targets. Matching FOV to site geometry is essential in perimeter and crowd monitoring deployments.

Flash LiDAR

Flash LiDAR illuminates an entire scene with a single wide laser pulse and captures the returning light across a detector array, much as a camera flash lights a whole room at once. Because nothing is steered there are no moving parts, and the full frame is captured simultaneously, avoiding the motion distortion that scanning architectures can introduce with fast-moving objects. The trade-off is range and resolution. Spreading laser energy across the whole field of view rather than concentrating it in a narrow beam limits detection distance, so flash designs are most common in short-range applications.

LiDAR vs. Camera

LiDAR and cameras are both used for object detection, but they capture fundamentally different data. LiDAR measures distance and produces precise 3D point clouds without capturing image data, making it inherently privacy-safe. Cameras capture rich visual detail but require image processing to extract depth, perform poorly in darkness or glare, and raise privacy compliance concerns. For applications requiring anonymous people tracking or 24/7 outdoor operation, 3D LiDAR offers advantages over camera-based systems. Quanergy’s Q-TRACK platform uses LiDAR to track people and vehicles without cameras or biometric data.

LiDAR vs. Radar

LiDAR uses laser pulses (typically near-infrared) to produce dense 3D point clouds with centimeter-level spatial resolution, enabling precise shape recognition and classification of people versus vehicles versus objects. Radar uses radio waves, offering longer range and better penetration through rain and fog, but with much lower spatial resolution and no shape data. For perimeter security applications where accurate object classification and low false alarm rates are priorities, 3D LiDAR typically outperforms radar. Radar performance also degrades at sites with significant metal infrastructure, because radio waves reflect strongly off fences, gates, shipping containers, parked vehicles, and metal-clad buildings, producing multipath returns and ghost targets that can be reported as intrusions. Radar is further susceptible to electromagnetic interference from nearby RF emitters, industrial equipment, and other radar units operating in proximity, which becomes a practical constraint when several units are required to cover a long perimeter. Because LiDAR operates at optical wavelengths rather than radio frequencies, it is unaffected by RF electromagnetic interference and does not generate the same multipath artifacts from metal surfaces. Radar remains preferred where extreme range or weather penetration is the primary requirement.

---

I

Indirect Time-of-Flight (iToF)

Indirect time-of-flight measures distance by emitting continuously modulated light and measuring the phase shift between the emitted and returning signal, rather than timing a discrete pulse. iToF produces dense, camera-like depth images and suits short and medium range applications where spatial detail matters more than reach. The Quanergy Q-VISION F540-W is a solid-state iToF sensor built for industrial automation and robotics, distinct from the direct time-of-flight ranging used in Q-TRACK.

IP Rating (Ingress Protection)

IP rating is an IEC 60529 standard that classifies the degree of protection a device enclosure provides against dust and water. The first digit (0 to 6) indicates solid particle protection; the second digit (0 to 9K) indicates liquid protection. IP66 means fully dust-tight and protected against powerful water jets. IP67 adds submersion protection up to one meter. Outdoor LiDAR sensors for perimeter security and infrastructure monitoring typically require a minimum of IP66 to operate reliably in rain, dust, and high-humidity environments.

---

L

Laser Channels

Laser channels (also called beams or rings) refer to the number of individual laser emitters in a LiDAR sensor, each firing at a different vertical angle. A sensor with more channels produces a denser vertical point distribution, improving the ability to detect low-profile objects like crouching intruders or small ground vehicles. Common channel counts range from 8 to 128 in multi-channel sensors. A higher channel count is not automatically a better solution. Every additional channel increases the volume of point cloud data the sensor must transmit and the system must store and process, which raises network bandwidth requirements and the specification of the supporting switching, cabling, and server infrastructure. Quanergy’s Q-TRACK sensors are designed with channel configurations optimized to detect and classify objects reliably at long range rather than to maximize raw channel count. This keeps per-sensor throughput requirements low and reduces the network and hardware burden of scaling to a large multi-sensor deployment.

LiDAR (Light Detection and Ranging)

LiDAR is a remote sensing technology that measures distance by emitting laser pulses and timing their return after reflecting off surfaces. A 3D LiDAR sensor fires thousands to millions of pulses per second across a defined field of view, building a three-dimensional map of its surroundings. Unlike cameras, LiDAR is not affected by ambient lighting conditions, works equally in daylight and complete darkness, and does not capture identifiable images. Quanergy manufactures 3D LiDAR sensors and software platforms used in security, crowd analytics, airport passenger flow, and industrial automation.

---

M

Mechanical Scanning LiDAR

Mechanical scanning LiDAR steers the laser by physically rotating the optical assembly, sweeping the beam through the environment to build a complete point cloud. A rotating assembly delivers continuous 360 degree horizontal coverage from a single sensor, which fixed-aperture architectures cannot match because they cover a limited angular sector. The trade-off is the presence of moving parts. For wide-area perimeter and crowd monitoring the coverage advantage is usually decisive, because one sensor surveys a large area continuously rather than looking through a narrow window. Quanergy Q-TRACK sensors (LR, HD, and Dome) and the Quanergy M-SERIES use mechanical scanning architectures. Q-TRACK LR covers a 70 metre radius, approximately 15,000 square metres per sensor.

MEMS LiDAR

MEMS LiDAR (Micro-Electro-Mechanical Systems) steers the laser using a microscopic mirror etched onto a silicon chip. The mirror oscillates at high frequency to sweep the beam across the field of view without rotating the sensor housing. MEMS designs can be compact and support programmable scan patterns. They are frequently marketed as solid-state, but the mirror is a physically moving component, so hybrid solid-state is the more precise description. MEMS sensors typically cover a limited angular sector rather than a full 360 degrees. Quanergy sensors do not use MEMS beam steering.

Multi-Object Tracking (MOT)

Multi-object tracking is the process of simultaneously maintaining identification, position, and trajectory data for multiple independent objects detected in a sensor’s field of view. A MOT system assigns each detected object a unique track ID, predicts its future position between sensor frames, and updates the track as new detections arrive. Quanergy Q-TRACK is capable of tracking more than 1,000 objects simultaneously, making it applicable in high-density crowd environments such as airport terminals and transit hubs.

Multi-Return

Multi-return capability allows a LiDAR sensor to record more than one return signal from a single laser pulse. This occurs when a pulse partially penetrates a semi-transparent object (such as foliage or chain-link fence) and reflects from both the near surface and a surface behind it. Multi-return processing allows the sensor to see through vegetation and wire mesh, improving detection reliability in cluttered outdoor environments commonly found at perimeter security sites.

---

O

Object Classification

Object classification is the process of categorizing a detected object by type based on its measured 3D shape, size, and movement characteristics. In LiDAR-based security systems, the primary classification categories are person, vehicle (with sub-types such as car, truck, and bicycle), and unknown object. Accurate object classification is essential for reducing nuisance alarms and for triggering the correct response protocol. 3D LiDAR enables more reliable object classification than 2D sensors because shape data is available in all three dimensions. Quanergy Q-TRACK uses shape-based object classification to achieve up to 95% reduction in false and nuisance alarms in perimeter security deployments.

Object Tracking

Object tracking is the continuous measurement of a detected object’s position and trajectory over time. Once an object is detected and classified, a tracking algorithm assigns it a unique ID and updates its location with each new sensor scan. Tracking persists even when the object is temporarily obscured by another object or moves out of one sensor’s view and into another’s. Quanergy Q-TRACK performs multi-object tracking for people and vehicles across distributed sensor networks with a claimed accuracy of up to 99%.

OPA LiDAR (Optical Phased Array)

OPA LiDAR steers the laser beam electronically rather than mechanically. An array of optical emitters manipulates the phase of emitted light so the wavefronts reinforce in a chosen direction, allowing the beam to be pointed with no moving parts of any kind. This makes OPA a true solid-state architecture, distinct from MEMS designs, which are often marketed as solid-state but still contain a microscopically moving mirror. OPA remains an active area of semiconductor photonics research and commercial deployments are limited. Quanergy’s current sensor portfolio does not use OPA. Quanergy 3D LiDAR sensors use mechanical scanning architectures, and the Q-VISION F540-W uses solid-state iToF.

---

P

Point Cloud

A point cloud is the set of 3D data points produced by a LiDAR sensor during a single scan cycle, each representing the location of a surface reflection in 3D space (X, Y, Z coordinates). A high-density point cloud from a multi-channel sensor may contain hundreds of thousands to millions of points per second. Point clouds are processed by detection and tracking algorithms to identify objects, measure their dimensions, and follow their movement. The raw point cloud data does not contain image or biometric information, making it inherently privacy-safe for people counting and tracking applications.

Point Density

Point density refers to the number of LiDAR points returned per unit area or per unit solid angle. Higher point density means more data points per object, improving detection of small targets and more accurate shape reconstruction for classification. Point density depends on the number of laser channels, the scan speed, and the range to the target. At longer ranges, point density decreases as the same number of beams covers a larger area.

Power over Ethernet (PoE)

Power over Ethernet is a networking standard (IEEE 802.3af, 802.3at, or 802.3bt) that delivers DC power over standard Ethernet cabling, eliminating the need for a separate power supply cable. PoE-capable LiDAR sensors require only a single Ethernet cable for both power and data, simplifying installation and reducing infrastructure cost in perimeter and building security deployments. Quanergy’s M1 Edge LiDAR sensor supports PoE operation, enabling single-cable installation in facilities where running separate power lines is impractical.

Pulse Repetition Rate

Pulse repetition rate (PRR) is the number of laser pulses a LiDAR sensor fires per second, expressed in pulses per second (PPS) or kilohertz. Higher pulse repetition rates produce more data points per second and can support faster scan rates or higher point density. However, very high repetition rates may cause range ambiguity if a pulse is fired before the previous one has returned from the maximum detection range.

---

R

Range Resolution

Range resolution is the minimum difference in distance between two objects along the LiDAR line of sight that the sensor can distinguish as separate targets. It is determined by the pulse width of the laser and the timing precision of the receiver electronics. Fine range resolution allows the sensor to separate objects that are close together in depth, such as a person standing near a wall or two pedestrians walking closely together.

Reflectivity

Reflectivity describes how much of the incident laser energy a surface returns to the LiDAR sensor. High-reflectivity surfaces (such as white paint or retroreflective tape) return strong signals and are detectable at long range. Low-reflectivity surfaces (such as dark clothing or wet asphalt) absorb more energy and may only be detectable at closer ranges. LiDAR sensor specifications typically state maximum detection range at a given target reflectivity, such as 10% (a challenging dark surface) or 80% (a bright surface).

---

S

Semantic Segmentation

Semantic segmentation in 3D LiDAR processing assigns a class label (such as ground, wall, person, or vehicle) to each individual point in a point cloud. Unlike object detection, which identifies bounding boxes around discrete objects, semantic segmentation classifies every point, enabling the system to understand complex scenes where objects overlap or blend into their environment. Semantic segmentation is used in advanced perception systems for autonomous vehicles and industrial automation where scene understanding must be complete and precise.

SLAM (Simultaneous Localization and Mapping)

SLAM is an algorithm category that allows a sensor-equipped device to build a map of an unknown environment while simultaneously tracking its own position within that map. In LiDAR-based SLAM, point clouds are matched between successive scans to estimate sensor motion and accumulate a 3D map. SLAM is widely used in mobile robotics, autonomous vehicles, and indoor positioning systems where GPS is unavailable.

Signal-to-Noise Ratio (SNR)

Signal-to-noise ratio is the ratio of the useful return signal power to background noise power in a LiDAR receiver. Higher SNR means cleaner return signals and more reliable range measurements. SNR degrades in high-ambient-light conditions (such as direct sunlight), rain, or fog, which can scatter the laser pulse and reduce the usable signal. LiDAR sensors designed for outdoor security use are engineered with optical filters and receiver designs that maximize SNR under adverse environmental conditions.

Solid-State LiDAR

Solid-state LiDAR refers to sensor designs that contain no rotating or mechanically oscillating components. Instead, they use electronic or micro-mechanical beam steering (such as MEMS approaches) or fixed-beam flash illumination. Solid-state designs are potentially more durable, smaller, and lower in cost than mechanical spinning LiDAR sensors, because they have fewer wear components. The trade-off is typically in field of view, as solid-state sensors often cover a limited angular sector rather than a full 360 degrees. Quanergy Q-VISION is a solid-state LiDAR sensor designed for close- to mid-range detection and classification in security and automation applications. Solid-state is an umbrella term rather than a single technology. OPA and flash designs are true solid-state, while MEMS is often marketed as solid-state despite containing a moving mirror.

SPAD (Single-Photon Avalanche Diode)

A SPAD is a semiconductor photodetector sensitive enough to register a single returning photon. It describes how a LiDAR sensor detects light, not how it steers the beam or measures distance, which is why SPAD is not an alternative to MEMS, OPA, or mechanical scanning. A sensor can pair a SPAD detector with any beam steering method. High detector sensitivity improves performance against low-reflectivity targets such as dark clothing or wet asphalt, and at longer ranges where the returning signal is weak.

---

T

Time-of-Flight (ToF)

Time-of-flight is the fundamental measurement principle used by LiDAR sensors. The sensor emits a laser pulse and records the elapsed time until the reflected pulse returns. Because light travels at a known constant speed (approximately 3 x 10^8 meters per second), the round-trip travel time directly yields the distance to the reflecting surface. A 3D LiDAR sensor repeats this measurement across many angles and channels to build a full point cloud of its surroundings. Time-of-flight is implemented in two distinct ways. Direct time-of-flight (dToF) times a discrete laser pulse, while indirect time-of-flight (iToF) measures the phase shift of continuously modulated light. Quanergy uses both approaches across its portfolio.

---

V

Voxel

A voxel (volumetric pixel) is the 3D equivalent of a pixel in a 2D image: a discrete cubic unit of volume in a 3D grid. Point cloud data can be converted to a voxel representation by assigning each point to the voxel containing it. Voxelization reduces data density for faster processing and enables spatial operations such as occupancy mapping and change detection. In crowd monitoring, voxel-based occupancy grids are used to measure crowd density across large areas.

---