Zhen Dong

Wuhan University | WHU · State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing

Point cloud semantic segmentation (PCSS) shows great potential in generating accurate 3D semantic maps for digital twin railways. Deep learning-based methods have seen substantial advancements, driven by numerous PCSS datasets. Nevertheless, existing datasets tend to neglect railway scenes, with limitations in scale, categories, and scene diversity...

Quick and high-precision semantic segmentation from Mobile Laser Scanning (MLS) point clouds faces huge challenges such as large amounts of data, occlusion in complex scenes, and the high annotation cost associated with 3D point clouds. To tackle these challenges, this paper proposes a novel efficient and high-precision semantic segmentation method...

OpenStreetMap (OSM), an online and versatile source of volunteered geographic information (VGI), is widely used for human self-localization by matching nearby visual observations with vectorized map data. However, due to the divergence in modalities and views, image-to-OSM (I2O) matching and localization remain challenging for robots, preventing th...

Wearable laser scanning (WLS) system has the advantages of flexibility and portability. It can be used for determining the user's path within a prior map, which is a huge demand for applications in pedestrian navigation, collaborative mapping, augmented reality, and emergency rescue. However, existing LiDAR-based global localization methods suffer...

Image-to-point cloud (I2P) registration is a fundamental task for robots and autonomous vehicles to achieve cross-modality data fusion and localization. Current I2P registration methods primarily focus on estimating correspondences at the point or pixel level, often neglecting global alignment. As a result, I2P matching can easily converge to a loc...

Due to the subterranean scene’s poor lighting conditions and variability of the environment, real-time localizing and meshing in complex underground scenes present a challenging task, with high-potential applications in mining and tunnel protection. In this work, we propose a method that combines SLAM and NeRF for mesh reconstruction in the undergr...

Dashboard cameras (dashcams) record millions of driving videos daily, offering a valuable potential data source for various applications, including driving map production and updates. A necessary step for utilizing these dashcam data involves the estimation of camera poses. However, the low-quality images captured by dashcams, characterized by moti...

Leveraging multi‐platform laser scanning systems offers a complete solution for 3D modelling of large‐scale urban scenes. However, the spatial inconsistency of point clouds collected by heterogeneous platforms with different viewpoints presents challenges in achieving seamless fusion. To tackle this challenge, this paper proposes a coarse‐to‐fine a...

Corresponding author}In this paper, we explore a novel framework, EGIInet (Explicitly Guided Information Interaction Network), a model for View-guided Point cloud Completion (ViPC) task, which aims to restore a complete point cloud from a partial one with a single view image. In comparison with previous methods that relied on the global semantics o...

Accurate estimation of above-ground biomass (AGB) in forested areas is essential for studying forest ecological functions, surface carbon cycling, and global carbon balance. Over the past decade, models that harness the distinct features of multi-source remote sensing observations for estimating AGB have gained significant popularity. It is worth e...

We propose SparseDC, a model for Depth Completion from Sparse and non-uniform inputs. Unlike previous methods focusing on completing fixed distributions on benchmark datasets (e.g., NYU with 500 points, KITTI with 64 lines), SparseDC is specifically designed to handle depth maps with poor quality in real usage. Our SparseDC makes two major contribu...

Precise and rapid delineation of sharp boundaries and robust semantics is essential for numerous downstream robotic tasks, such as robot grasping and manipulation, real-time semantic mapping, and online sensor calibration performed on edge computing units. Although boundary detection and semantic segmentation are complementary tasks, most studies f...

面向国家“双碳”目标和国际碳交易市场需求，陆地生态系统的固碳现状和未来固碳潜力亟须研究。森林是陆地生态系统中重要的碳库，目前基于地面观测的清查方法工作量大且抽样统计结果难以评价，基于卫星遥感反演的方法缺乏理论解释且普适性差。本文从单木级森林碳储量模型出发，提出一种基于遥感的森林碳储量显式计算模型。首先使用图像分辨率、森林郁闭度和森林高度3个关键变量构建森林碳储量显式计算模型，并对模型变量和参数进行误差分析；然后利用单木的生长特性，仿真不同饱和度的森林场景，从理论上解算不同树种的模型参数，并检验模型参数的精度与稳定性；最后验证模型在不同空间尺度、饱和度森林场景下的精度、稳健性和适用性。本文提出的森林碳储量模型解决了现有卫星遥感反演缺乏理论解释性和适用性差的难题，可实现大范围森林碳储量高分辨率制...

Assessing the solar photovoltaic (PV) potential on buildings is essential for environmental protection and sustainable development. However, currently, the high costs of data acquisition and labor required to obtain 3D building models limit the scalability of such estimations extending to a large scale. To overcome the limitations, this study propo...

Cross-source (CS) point cloud registration is a prerequisite for effectively leveraging the complementary information of multiple 3-D sensors. However, existing point cloud registration methods have primarily focused on the registration of mono-source point clouds and typically fail to register CS data with varying noise patterns and capture charac...

Point cloud semantic segmentation helps Intelligent Transportation Systems understand traffic scenes by assigning semantic label to each point in the point cloud, and it relies on large amounts of annotated training data. Nevertheless, manually annotating large-scale datasets of complex traffic scenes is quite time-consuming and tedious. This paper...

Point Cloud Place Recognition (PCPR) in street scenes is an essential task in the fields of autonomous driving, robot navigation, and urban map updating. However, the domain gap between heterogeneous point clouds and the difficulty of feature characterization in large-scale complex street scenes pose significant challenges for existing PCPR methods...

The position consistency of mobile laser scanning (MLS) point clouds is crucial for large-scale applications, and is normally guaranteed by the global navigation satellite system (GNSS) and high-precision inertial measurement unit (IMU) in the data acquisition. However, GNSS-denied environments such as city valleys result in significant position in...

Unpaired 3D object completion aims to predict a complete 3D shape from an incomplete input without knowing the correspondence between the complete and incomplete shapes. In this paper, we propose the novel KTNet to solve this task from the new perspective of knowledge transfer. KTNet elaborates a teacher-assistant-student network to establish multi...

Three-dimensional (3D) object detection utilizes numerous onboard sensors to determine the position, size, and motion information of surrounding objects. Recently, some researchers have utilized HD maps in 3D object detection for LiDAR point clouds. However, existing LiDAR–map fusion detection methods simply take the HD map as an additional input a...

Real-time 3-D mapping of large-scale global navigation satellite system (GNSS)-denied environments plays an important role in forest inventory management, disaster emergency response, and underground facility maintenance. Compact helmet laser scanning (HLS) systems keep the same direction as the user’s line of sight and have the advantage of “what...

Rigorous boresight calibration between light detection and ranging (LiDAR) and the camera is crucial for geometry and optical information fusion in earth observation and robotic applications. Although boresight parameters can be obtained through pre-calibration with artificial targets, unforeseen movement of sensors during data collection can lead...

With the rapid development of reality capture methods, such as laser scanning and oblique photogrammetry, point cloud data have become the third most important data source, after vector maps and imagery. Point cloud data also play an increasingly important role in scientific research and engineering in the fields of Earth science, spatial cognition...

Forest field inventory plays a crucial role in forestry management and the estimation of carbon circular economy, as it provides information on forest parameters, assesses carbon storage, and identifies the impact factors of ecological change. Terrestrial Laser Scanning (TLS) and Mobile Mapping Systems (MMS) are commonly used for forest field inven...

Wood-leaf separation from terrestrial laser scanning (TLS) is a crucial prerequisite for quantifying many biophysical properties and understanding ecological functions. In this study, we propose a novel multi-directional collaborative convolutional neural network (MDC-Net) that takes the original 3D coordinates and useful features from prior knowle...

In this paper, we present a new method for the multiview registration of point cloud. Previous multiview registration methods rely on exhaustive pairwise registration to construct a densely-connected pose graph and apply Iteratively Reweighted Least Square (IRLS) on the pose graph to compute the scan poses. However, constructing a densely-connected...

With the rapid expansion of urban areas in both horizontal and vertical directions, the complicated building structural changes challenge the existing 3D change detection methods. The existing 3D change detection methods are mainly based on local differences and rely on setting thresholds and rules, and face difficulties when determining complex ch...

We present RoReg, a novel point cloud registration framework that fully exploits oriented descriptors and estimated local rotations in the whole registration pipeline. Previous methods mainly focus on extracting rotation-invariant descriptors for registration but unanimously neglect the orientations of descriptors. In this paper, we show that the o...

Unpaired 3D object completion aims to predict a complete 3D shape from an incomplete input without knowing the correspondence between the complete and incomplete shapes. In this paper, we propose the novel KT-Net to solve this task from the new perspective of knowledge transfer. KT-Net elaborates a teacher-assistant-student network to establish mul...

The number of visible global navigation satellite system (GNSS) satellites is an important indicator for evaluating positioning accuracy. In urban areas, buildings and trees cause serious satellite signal obstruction and attenuation. Studies have used three-dimensional (3D) city models or 2D panoramic imagery to calculate the visibility of satellit...

Retrieval of glacier surface heights from ICESat-2 photon-counting data is significant to monitor glacier surface morphologies (e.g., crevasses) and their changes. However, existing methods are susceptible to complex glacier surfaces and diverse signal-to-noise ratios (SNRs). Therefore, we propose a robust density estimation method for glacier-heig...

Terrestrial laser scanning (TLS) is an important means to monitor landslides, and the layout is the key to guarantee captured point clouds with a high quality and low cost. Nevertheless, TLS layout in landslide monitoring is currently determined by user’s subjective experience, and shows a poor performance in reliability and point accuracy. Therefo...

Planning a terrestrial laser scanning (TLS) observation network is vital for capturing large-scale building facade point clouds with high time efficiency and high quality. However, existing planning methods are susceptible to low reliability, computational burden and lack of robust registrability assessment solutions, especially for a large-scale t...

The Increasing availabilities of Terrestrial Laser Scanning (TLS) and Unmanned Aerial Vehicle Laser Scanning (ULS) have advanced accurate and detailed forest structural measurements. Registration of multi-perspective observations from different platforms is a prerequisite for a comprehensive forest structure understanding. Currently, forest point c...

Benefiting from the development of deep learning, researchers have made significant progress and achieved superior performance in the semantic segmentation of remote sensing (RS) data. However, when encountering an unseen scenario, the performance of a trained model deteriorates dramatically because of the domain shift. Unsupervised domain adaptati...

Given the initial calibration of multiple sensors, the fine registration between Mobile Laser Scanning (MLS) point clouds and panoramic images is still challenging due to the unforeseen movement and temporal misalignment while collecting. To tackle this issue, we proposed a novel automatic method to register the panoramic images and MLS point cloud...

Point cloud upsampling is to densify a sparse point set acquired from 3D sensors, providing a denser representation for the underlying surface. Existing methods divide the input points into small patches and upsample each patch separately, however, ignoring the global spatial consistency between patches. In this paper, we present a novel method PC2...

Outdated or sketchy inventory of street furniture may misguide the planners on the renovation and upgrade of transportation infrastructures, thus posing potential threats to traffic safety. Previous studies have taken their steps using point clouds or street-view imagery (SVI) for street furniture inventory, but there remains a gap to balance seman...

Point cloud shape completion plays a central role in diverse 3D vision and robotics applications. Early methods used to generate global shapes without local detail refinement. Current methods tend to leverage local features to preserve the observed geometric details. However, they usually adopt the convolutional architecture over the incomplete poi...

Reconstructing 3D geometry from \emph{unoriented} point clouds can benefit many downstream tasks. Recent methods mostly adopt a neural shape representation with a neural network to represent a signed distance field and fit the point cloud with an unsigned supervision. However, we observe that using unsigned supervision may cause severe ambiguities...

Point cloud-based place recognition is a fundamental part of the localization task, and it can be achieved through a retrieval process. Reranking is a critical step in improving the retrieval accuracy, yet little effort has been devoted to reranking in point cloud retrieval. In this paper, we investigate the versatility of rigid registration in rer...

目前全景图像位置和姿态参数的解算多基于点特征，而场景中普遍存在的线特征尚未得到充分利用。本文提出一种点-线特征联合的全景图像位姿解算方法，不仅可用于点特征缺失场景中全景图像位姿参数的解算，而且在点特征充足的场景中可提高位姿解算的精度和稳健性。该方法中的线特征使用线上的任意两点表示，不要求全景图像和三维场景同名线上的选点具有对应关系，因而易于选取，具有极大的实用性。首先，使用直接线性变换(DLT)构建点-线特征联合的全景图像位姿解算模型，并针对水平线和垂直线获取简化后的模型；然后，利用仿真道路场景，从特征点和线的不同组合方式以及大姿态角两方面分析该模型 的适用性，并通过人工引入不同类型及量级的点-线误差分析该模型的容差性；最后，将本文方法应用于全景图像与激光点云的融合，从理论和实践两方面证明点...

Line segment matching is important in applications that require recovering the 3D structure of objects (e.g., manmade objects in street-level scenarios). However, differentiating between true and false line matches is generally difficult without strong geometric constraints for line segments. Hence, additional constraints are forced to be used, sac...

Individual tree segmentation in forest scenes provides a foundation for forest ecosystem modeling and biodiversity assessment applications. Existing approaches work well for the cases where trees do not grow in layers. However, they may fail in the scenario with understory vegetation occlusion and heavily overlapped crowns. In this work, we propose...

Registration of large-scale outdoor Terrestrial Laser Scanning (TLS) point clouds remains many challenges in the scenes with symmetric and repetitive elements (e.g., park, forest, and tunnel), the weak geometric features (e.g., underground excavation), and dramatically changes in different phases (e.g., mountain). To address these issues, a novel n...

To register mobile mapping system (MMS) lidar points and panoramic-image sequences, a relative orientation model of panoramic images (PROM) is proposed. The PROM is suitable for cases in which attitude or orientation parameters are unknown in the panoramic-image sequence. First, feature points are extracted and matched from panoramic-image pairs us...

Unpaired 3D object completion aims to predict a complete 3D shape from an incomplete input without knowing the correspondence between the complete and incomplete shapes during training. To build the correspondence between two data modalities, previous methods usually apply adversarial training to match the global shape features extracted by the enc...

The question of whether each building of housing estate has equal access to nearby social service resources (e.g., public transportation service, catering, entertainment, etc.) is a major concern of citizens. This paper takes Wuhan as a case to explore the equality in social service resource sharing of the housing estate at a microscopic level by a...

Roadside trees are an important component of the urban ecosystem, and extracting their location based on a point cloud obtained through mobile laser scanning (MLS) is essential for the urban ecology, but remains challenging because of heavy occlusions, disturbance due to tree stake systems, and overlaps between furniture on the street and the crown...

Road markings are of great significance to road inventory management, intelligent transportation systems, high-definition maps (HD Maps), and autonomous driving. Most existing methods focus on extracting and classifying the road markings from mobile laser scanning (MLS) point clouds. Nevertheless, the performance suffers from the wear and incomplet...

Point cloud upsampling is to densify a sparse point set acquired from 3D sensors, providing a denser representation for underlying surface. However, existing methods perform upsampling on a single patch, ignoring the coherence and relation of the entire surface, thus limiting the upsampled capability. Also, they mainly focus on a clean input, thus...

In this paper, we propose a novel local descriptor-based framework, called You Only Hypothesize Once (YOHO), for the registration of two unaligned point clouds. In contrast to most existing local descriptors which rely on a fragile local reference frame to gain rotation invariance, the proposed descriptor achieves the rotation invariance by recent...