본 자료는 자율주행, 자율 주행, LIDAR, 도로 지도 제작, 차량 관리, 내비게이션 알고리즘 및 경로 계획 등 다양한 주제를 다루는 전 세계 영어 논문 및 특허를 엄선하여 정리한 최신 자료입니다. 여러 온라인 과학 저널에 게재된 자료들을 포함하고 있습니다.
Light detection and ranging (lidar) system and operating method thereof
Patent published on the 2026-05-21 in US under Ref US20260140240 by SAMSUNG ELECTRONICS CO LTD [KR] (Park Sangyun [kr], You Jangwoo [kr], Choi Jinwoo [kr], Ko Byunghoon [kr], Kim Sunil [kr])
Abstract: A light-detection and ranging (LiDAR) system includes a signal generator configured to generate a plurality of lights each having a different wavelength from each other, a transceiver including a transmitter configured to emit the plurality of lights as a transmission light and a receiver configured to receive a reception light obtained by reflection of the transmission light from a target object, an optical path separator in the transceiver and configured to separate an optical path of the tran[...]
Our summary: A LiDAR system utilizes a signal generator to produce multiple wavelengths of light. It includes a transceiver that emits and receives light reflected from target objects. An optical path separator is employed to distinguish between the transmission and reception light paths.
LiDAR, signal generator, transceiver, optical path separator
Patent
Parking control methods, electronic devices, and storage media
Patent published on the 2026-05-21 in WO under Ref WO2026103377 by ZHEJIANG HUARAY TECH CO LTD [CN] (Hu Kun [cn], Shen Pengjie [cn], Zhao Ming [cn], Hu Lizhi [cn], Lu Wei [cn])
Abstract: A parking control method, electronic device, and storage media. The method comprises the following operations. A position of a reflective target is determined by a LiDAR. A vehicle is controlled to move to a recognition area of a visual marker. A pose relationship between the vehicle and the visual marker is determined based on a visual sensor within the recognition area of the visual marker. A parking target point is determined based on the pose relationship. The vehicle is controlled to move t[...]
Our summary: The method utilizes LiDAR to determine the position of a reflective target. A visual sensor assesses the pose relationship between the vehicle and a visual marker. The vehicle is then guided to a designated parking target point based on this information.
LiDAR, visual sensor, parking control, vehicle navigation
Patent
Accurate gas leak detection using tuneable diode lidar systems
Patent published on the 2026-05-21 in US under Ref US20260140258 by QLM TECH LTD [GB] (Ai Xiao [gb], Titchener James Graham [gb], Reed Murray Keith [gb])
Abstract: When using spectroscopic diode lidar to detect a gas of interest such as methane (CH4), detecting high concentration pathlength (CPL) values of the gas of interest may not actually indicate a plume or leak. In areas such as a refinery, where ambient gas levels may be routinely high, absorption from ambient gas measured over a long pathlength may produce the same CPL value as a much closer gas leak, leading to a false positive. By collecting and using lidar distance information in tandem with the[...]
Our summary: Tuneable diode lidar systems enhance gas leak detection accuracy. High concentration pathlength values can mislead detection in high ambient gas environments. Integrating lidar distance data with CPL values improves reliability and alarm triggering for gas leaks.
lidar, gas detection, methane, spectroscopic measurements
Patent
Autonomous uav visual navigation system
Patent published on the 2026-05-21 in US under Ref US20260140503 by SAUDI DATA AND ARTIFICIAL INTELLIGENCE AUTHORITY SDAIA [SA] (Samma Hussein Salem Ali Bin [sa], Elferik Sami [sa])
Abstract: [0000] A system, method, and non-transitory computer readable medium for autonomous vehicle navigation and visual inspection utilizing a deep Q-network and reinforcement learning includes at least one camera mounted to the vehicle for capturing continuous image frames of a scene, and processing circuitry configured with a reinforcement learning engine that generates a next control action for vehicle movement based on the captured image frames and a reward for a previous control action. A self-su[...]
Our summary: The system utilizes a deep Q-network and reinforcement learning for autonomous navigation and visual inspection. It captures continuous image frames with a mounted camera and processes them to determine control actions. A remote display terminal communicates the captured images and controls UAV movement to avoid obstacles.
autonomous navigation, deep Q-network, reinforcement learning, UAV
Patent
Autonomous vehicle and autonomous driving method for vehicle
Patent published on the 2026-05-21 in US under Ref US20260138634 by KIA CORP [KR] (Nam Yun Hyeok [kr], Lee Myung Soo [kr], Kim Hyeon Ji [kr], Shin Sang Ho [kr], Kim Chan Ryeol [kr])
Abstract: [0000] An apparatus of a vehicle may comprise a processor and a memory storing at least one instruction that, when executed by the processor communicating with the memory, is configured to cause the apparatus to identify a potential target within a threshold distance from the vehicle, determine, based on the potential target, whether the vehicle is in a dangerous situation, determine, based on data from a sensor of the vehicle, whether a driver of the vehicle is careless, and, based on determini[...]
Our summary: The vehicle apparatus includes a processor and memory for identifying potential targets. It assesses dangerous situations based on sensor data and driver behavior. If a danger is detected and the driver is careless, it activates a warning signal.
autonomous vehicle, driving method, sensor data, warning system
Patent
Computer system and method for obstacle perception
Patent published on the 2026-05-20 in EP under Ref EP4745620 by VOLVO AUTONOMOUS SOLUTIONS AB [SE] (Gustavsson Viktor [se])
Abstract: [0001] A computer system comprising processing circuitry is provided. The processing circuitry is configured to: obtain a first local ground plane estimation (LE1) of a first ground area (GA1) based on a first set of LIDAR data; obtain a second local ground plane estimation (LE2) of a second ground area (GA2), at least partly overlapping the first ground area (GA1), based on a second set of LIDAR data; fuse the first and second local ground plane estimations (LE1, LE2) to form a fused ground pla[...]
Our summary: The system processes LIDAR data to estimate local ground planes in overlapping areas. It fuses these estimations to create a comprehensive ground plane. The fused data is then used to identify obstacles above the ground.
LIDAR, obstacle perception, ground plane estimation, data fusion
Patent
Modeling and Correction of Underwater Photon-Counting LiDAR Returns Based on a Modified Biexponential Distribution
Published on 2026-02-03 by Jie Wang, Wei Hao, Songmao Chen, Meilin Xie, Heng Shi, Xiangyu Li, Xuezheng Lian, Xiuqin Su, Runqiang Xing, Lu Ding @MDPI
Abstract: Laser pulses experience significant temporal broadening in underwater environments due to strong turbulence and scattering effects. As water turbidity increases, the likelihood of multiple scattering events rises, further intensifying pulse broadening and thereby degrading the ranging accuracy of underwater single-photon LiDAR systems. Accurate characterization of the return pulse shape is crucial for precise distance extraction, typically achieved via cross-correlation with the system&r[...]
Our summary: This study introduces a Modified Biexponential Distribution model to accurately characterize the return pulse shape of underwater LiDAR systems. The model addresses the issues of pulse broadening caused by water turbidity and scattering, improving ranging accuracy. Experimental results show a significant reduction in Depth Absolute Error, confirming the model s effectiveness in enhancing underwater photon-counting LiDAR performance.
Photon-Counting, LiDAR, Biexponential Distribution, Underwater
Publication
Hybrid Mamba–Graph Fusion with Multi-Stage Pseudo-Label Refinement for Semi-Supervised Hyperspectral–LiDAR Classification
Published on 2026-02-03 by Khanzada Muzammil Hussain, Keyun Zhao, Sachal Perviaz, Ying Li @MDPI
Abstract: Semi-supervised joint classification of Hyperspectral Images (HSIs) and LiDAR-derived Digital Surface Models (DSMs) remains challenging due to scarcity of labeled pixels, strong intra-class variability, and the heterogeneous nature of spectral and elevation features. In this work, we propose a Hybrid Mamba–Graph Fusion Network (HMGF-Net) with Multi-Stage Pseudo-Label Refinement (MS-PLR) for semi-supervised hyperspectral–LiDAR classification. The framework employs a sp[...]
Our summary: This work presents a Hybrid Mamba-Graph Fusion Network (HMGF-Net) for semi-supervised hyperspectral-LiDAR classification. The framework integrates spectral-spatial HSI backbones and LiDAR CNN encoders with a graph fusion module for enhanced information propagation. Validation on benchmark datasets shows significant performance improvements over state-of-the-art methods, particularly in low-label scenarios.
Hybrid Mamba-Graph Fusion, Semi-Supervised Learning, Hyperspectral Classification, LiDAR Integration
Publication