A5067350621- Robotics and Sensor-Based Localization
- Robotic Path Planning Algorithms
- Indoor and Outdoor Localization Technologies
- Advanced Vision and Imaging
- Remote Sensing and LiDAR Applications
- Advanced Optical Sensing Technologies
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Inertial Sensor and Navigation
- Infrastructure Maintenance and Monitoring
- CCD and CMOS Imaging Sensors
- Advanced Neural Network Applications
- Robotic Mechanisms and Dynamics
- Asphalt Pavement Performance Evaluation
- Image Processing Techniques and Applications
- Advanced Memory and Neural Computing
- Vehicular Ad Hoc Networks (VANETs)
- Optical measurement and interference techniques
- Ruminant Nutrition and Digestive Physiology
- Age of Information Optimization
- Autonomous Vehicle Technology and Safety
- Food Supply Chain Traceability
- Geotechnical Engineering and Underground Structures
- Advanced Manufacturing and Logistics Optimization
- Soft Robotics and Applications
Hong Kong University of Science and Technology
2007-2025
University of Hong Kong
2007-2025
Shanghai Fudan Microelectronics (China)
2024
Fudan University
2024
Gorgias Press (United States)
2021
Institute of Electrical and Electronics Engineers
2021
Stevens Institute of Technology
2021
China Agricultural University
2010-2016
Institute of Grassland Research
2016
Hefei University of Technology
2012
Crack is one of the most common road distresses which may pose safety hazards. Generally, crack detection performed by either certified inspectors or structural engineers. This task is, however, time-consuming, subjective and labor-intensive. In this paper, a novel algorithm based on deep learning adaptive image segmentation proposed. Firstly, convolutional neural network trained to determine whether an contains cracks not. The images containing are then smoothed using bilateral filtering,...
Combining multiple LiDARs enables a robot to maximize its perceptual awareness of environments and obtain sufficient measurements, which is promising for simultaneous localization mapping (SLAM). This article proposes system achieve robust extrinsic calibration, odometry, LiDARs. Our approach starts with measurement preprocessing extract edge planar features from raw measurements. After motion initialization procedure, sliding window-based multi-LiDAR odometry runs onboard estimate poses an...
Multiple lidars are used on mobile vehicles for rendering a broad view to enhance the performance of perception systems. However, precise calibration multiple is challenging since feature correspondences in scan points sparse providing enough constraints. To address this problem, existing methods require fixed targets scenes or rely exclusively additional sensors. In paper, we present novel method that enables automatic lidar without these restrictions. Three linearly independent planar...
Modern LiDAR-SLAM (L-SLAM) systems have shown excellent results in large-scale, real-world scenarios. However, they commonly a high latency due to the expensive data association and nonlinear optimization. This paper demonstrates that actively selecting subset of features significantly improves both accuracy efficiency an L-SLAM system. We formulate feature selection as combinatorial optimization problem under cardinality constraint preserve information matrix's spectral attributes. The...
Combining multiple sensors enables a robot to maximize its perceptual awareness of environments and enhance robustness external disturbance, crucial robotic navigation. This paper proposes the FusionPortable benchmark, complete multi-sensor dataset with diverse set sequences for mobile robots. presents three contributions. We first advance portable versatile suite that offers rich sensory measurements: 10Hz LiDAR point clouds, 20Hz stereo frame images, high-rate asynchronous events from...
This paper investigates two typical image-type representations for event camera-based tracking: time surface (TS) and map (EM). Based on the original TS-based tracker, we make use of these representations' complementary strengths to develop an enhanced version. The pro-posed tracker consists a general strategy evaluate optimization problem's degeneracy online then switch proper representations. Both TS EM are motion- scene-dependent, thus it is important figure out their limitations in...
Using different sensors in an autonomous vehicle (AV) can provide multiple constraints to optimize AV location estimation. In this paper, we present a low-cost GPS-assisted LiDAR state estimation system for AVs. Firstly, utilize obtain highly precise 3D geometry data. Next, use inertial measurement unit (IMU) correct point cloud misalignment caused by incorrect place recognition. The estimated odometry and IMU are then jointly optimized. We GPS instead of realtime kinematic (RTK) module...
Pose estimation problems are fundamental in robotics. Most of these challenging due to the nonconvex nature. This also sets up an obstacle for uncertainty description that is essential pose integration and quality control. In this article, we show a large class related can be categorized as quadratic (QPEPs) propose general quaternion-based mathematical model unify problems. To solve QPEPs, Gröbner-basis method investigated derive their globally optimal robust solutions. Furthermore, develop...
Multi-LiDAR systems have been prevalently applied in modern autonomous vehicles to render a broad view of the environments. The rapid development 5G wireless technologies has brought breakthrough for current cellular vehicle-to-everything (C-V2X) applications. Therefore, novel localization and perception system which multiple LiDARs are mounted around cities proposed. However, existing calibration methods require specific hard-to-move markers, ego-motion, or good initial values given by...
Simultaneous Localization and Mapping (SLAM) has been widely applied in various robotic missions, from rescue operations to autonomous driving. However, the generalization of SLAM algorithms remains a significant challenge, as current datasets often lack scalability terms platforms environments. To address this limitation, we present FusionPortableV2, multi-sensor dataset featuring sensor diversity, varied motion patterns, wide range environmental scenarios. Our comprises 27 sequences,...
Intelligent transportation systems have become increasingly important for the public in Shanghai. In response, ShanghaiGrid (SG) aims to provide abundant intelligent services improve traffic condition. A challenging service SG is accurately locate positions of moving vehicles real time. this paper we present an innovative scheme HERO tackle problem. SG, location information individual actively logged local nodes which are distributed throughout city. For each vehicle, dynamically maintains...
In this paper we address the problem of energy-efficient event detection in wireless sensor networks (WSNs). Duty cycling is a fundamental approach to conserving energy WSNs. However, it brings challenges sense that an may be undetected or undergo certain delay before detected, particular when sensors are low duty-cycled. We investigate relationship between and efficiency. Based on simplified network model, quantify performance by deriving closed forms detectability. also characterize...
Combining corn with legumes for silage is a feasible strategy to improve crude protein concentration in silage. This study was conducted determine the fermentation characteristics and overall nutritive value of ensiled (Zea mays L.), vine peas (Pisum sativum mixtures two crops at three different ratios. The were corn/pea containing 0, 33, 50, 67, 100% vacuumed plastic bags 60 days room temperature. addition dry matter concentrations forages had negligible influence on neutral detergent fibre...
Localization, or position fixing, is an important problem in robotics research. In this paper, we propose a novel approach for long-term localization changing environment using 3D LiDAR. We first create the map of real GPS and Then, divide into several small parts as targets cloud registration, which can not only improve robustness but also reduce registration time. proposed method called PointLocalization. PointLocalization allows us to fuse different kinds odometers, optimize accuracy...
Since early 2020, COVID-19 has spread rapidly across the world. At date of this writing, disease been globally reported in 220 countries and regions, infected more than 68 million people, caused 1.5 deaths (see https://covid19.who.int/) for up-to-date statistics. Avoiding person-to-person transmission is an effective approach to control prevent pandemic. However, many daily activities, such as transporting goods our life, inevitably involve contact.
Low-cost Inertial Measurement Units (IMUs) are ubiquitously used in the attitude estimation of cell phones and robots. Accurate robust IMU calibration is required to ensure accuracy. This paper proposes an accurate equipment-free method. We do not assume gyroscope biases invariant during compared with conventional methods because low-cost IMUs like MPU6000 have large instability. propose introducing removal outlier-aware optimization alleviate impact variant biases. Furthermore, we introduce...
Hand-eye calibration (HEC) is a kernel technique guaranteeing precision industrial visual servoing and robotic grasping. Extensive studies have been conducted to various closed-form iterative solutions HEC problems using different pose parameterizations. However, these approaches are either sensitive input noise or time-consuming for implementation. This article provides new perspective on deterministic solution two major branches of forms <b xmlns:mml="http://www.w3.org/1998/Math/MathML"...
In this paper, we present an autonomous unmanned aerial vehicle (UAV) landing system based on visual navigation. We design the landmark as a topological pattern in order to enable UAV distinguish from environment easily. addition, dynamic thresholding method is developed for image binarization improve detection efficiency. The relative distance horizontal plane calculated according effective information, and height obtained using linear interpolation method. experiments are performed static...
Combining multiple LiDARs enables a robot to maximize its perceptual awareness of environments and obtain sufficient measurements, which is promising for simultaneous localization mapping (SLAM). This paper proposes system achieve robust extrinsic calibration, odometry, LiDARs. Our approach starts with measurement preprocessing extract edge planar features from raw measurements. After motion initialization procedure, sliding window-based multi-LiDAR odometry runs onboard estimate poses...
Crack is one of the most common road distresses which may pose safety hazards. Generally, crack detection performed by either certified inspectors or structural engineers. This task is, however, time-consuming, subjective and labor-intensive. In this paper, we propose a novel algorithm based on deep learning adaptive image segmentation. Firstly, convolutional neural network trained to determine whether an contains cracks not. The images containing are then smoothed using bilateral filtering,...
The loop closure problem in 2D LiDAR simultaneous localization and mapping (SLAM) is interesting yet to be solved efficiently, as it suffers from a lack of information laser rangefinder readings. For this reason, we compare two grid submaps, where features constraints are enriched, find loops. We propose geometric environment descriptor called Triangle Feature (TF). It exploits the Euclidean distance constraint any three feature points submap can form. A graph SLAM system using TF close...