Review of Automatic Processing of Topography and Surface Feature Identification LiDAR Data Using Machine Learning Techniques

Gharineiat, Zahra; Kurdi, Fayez Tarsha; Campbell, Glenn

doi:10.3390/rs14194685

Cited by 28 publications

(21 citation statements)

References 148 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The idea is simple: if it is possible to accurately identify which points belong to pedestrians, vehicles, bicycles, or others then the dynamic point cloud can be easily filtered out. In recent years, machine learning techniques have been widely used in segmentation-based methods [ 67 , 68 ]. This method uses semantic segmentation based on deep learning to label the class of dynamic objects so that the dynamic point cloud can be eliminated directly with a bounding box, as shown in Figure 6 .…”

Section: Dynamic Target Filtering Methods In the 3d Point Cloudmentioning

confidence: 99%

A Review of Dynamic Object Filtering in SLAM Based on 3D LiDAR

Peng,

Zhao,

Wang

2024

Sensors

View full text Add to dashboard Cite

SLAM (Simultaneous Localization and Mapping) based on 3D LiDAR (Laser Detection and Ranging) is an expanding field of research with numerous applications in the areas of autonomous driving, mobile robotics, and UAVs (Unmanned Aerial Vehicles). However, in most real-world scenarios, dynamic objects can negatively impact the accuracy and robustness of SLAM. In recent years, the challenge of achieving optimal SLAM performance in dynamic environments has led to the emergence of various research efforts, but there has been relatively little relevant review. This work delves into the development process and current state of SLAM based on 3D LiDAR in dynamic environments. After analyzing the necessity and importance of filtering dynamic objects in SLAM, this paper is developed from two dimensions. At the solution-oriented level, mainstream methods of filtering dynamic targets in 3D point cloud are introduced in detail, such as the ray-tracing-based approach, the visibility-based approach, the segmentation-based approach, and others. Then, at the problem-oriented level, this paper classifies dynamic objects and summarizes the corresponding processing strategies for different categories in the SLAM framework, such as online real-time filtering, post-processing after the mapping, and Long-term SLAM. Finally, the development trends and research directions of dynamic object filtering in SLAM based on 3D LiDAR are discussed and predicted.

show abstract

Section: Dynamic Target Filtering Methods In the 3d Point Cloudmentioning

confidence: 99%

A Review of Dynamic Object Filtering in SLAM Based on 3D LiDAR

Peng,

Zhao,

Wang

2024

Sensors

View full text Add to dashboard Cite

show abstract

“…Recent advancements in remote sensing have widened the range of applications for 3D Point Cloud (PC) data. This data format poses several new issues concerning noise levels, sparsity and required storage space; as a result, many recent works address PC problems using deep learning solutions due to their capability to automatically extract features and achieve high performances [18,19].…”

Section: Introductionmentioning

confidence: 99%

Arbitrary-Oriented Object Detection in Aerial Images with Dynamic Deformable Convolution and Self-Normalizing Channel Attention

Zhang

et al. 2023

Electronics

View full text Add to dashboard Cite

Objects in aerial images often have arbitrary orientations and variable shapes and sizes. As a result, accurate and robust object detection in aerial images is a challenging problem. In this paper, an arbitrary-oriented object detection method for aerial images, based on Dynamic Deformable Convolution (DDC) and Self-normalizing Channel Attention Mechanism (SCAM), is proposed; this method uses ReResNet-50 as the backbone network to extract rotation-equivariant features. First, DDC is proposed as a replacement for the conventional convolution operation in the Convolutional Neural Network (CNN) in order to cope with various shapes, sizes and arbitrary orientations of the objects. Second, SCAM embedded into the high layer of ReResNet-50, which allows the network to enhance the important feature channels and suppress the irrelevant ones. Finally, Rotation Regions of Interest (RRoI) are generated based on a Region Proposal Network (RPN) and a RoI Transformer (RT), and the RoI-wise classification and bounding box regression are realized by Rotation-invariant RoI Align (RiRoI Align). The proposed method is comprehensively evaluated on three publicly available benchmark datasets. The mean Average Precision (mAP) can reach 80.91%, 92.73% and 94.1% on DOTA-v1.0, DOTA-v1.5 and HRSC2016 datasets, respectively. The experimental results show that, when compared with the state-of-the-arts methods, the proposed method can achieve superior detection accuracy.

show abstract

“…In the past ten years, deep learning has developed rapidly and made major breakthroughs. Related methods have been successfully applied to cameras and LiDARs [1][2][3]. However, due to the limitations of cameras and LiDARs, it is difficult to detect targets around the vehicle in harsh environments, so this paper refers to a new radar data format.…”

Section: Introductionmentioning

confidence: 99%

Radar Timing Range–Doppler Spectral Target Detection Based on Attention ConvLSTM in Traffic Scenes

Jia,

Tan,

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

With the development of autonomous driving and the emergence of various intelligent traffic scenarios, object detection technology based on deep learning is more and more widely applied to real traffic scenarios. Commonly used detection devices include LiDAR and cameras. Since the implementation of traffic scene target detection technology requires mass production, the advantages of millimeter-wave radar have emerged, such as low cost and no interference from the external environment. The performance of LiDAR and cameras is greatly reduced due to their sensitivity to light, which affects target detection at night and in bad weather. However, millimeter-wave radar can overcome the influence of these harsh environments and has a great auxiliary effect on safe driving on the road. In this work, we propose a deep-learning-based object detection method considering the radar range–Doppler spectrum in traffic scenarios. The algorithm uses YOLOv8 as the basic architecture, makes full use of the time series characteristics of range–Doppler spectrum data in traffic scenarios, introduces the ConvLSTM network, and exerts the ability to process time series data. In order to improve the model’s ability to detect small objects, an efficient and lightweight Efficient Channel Attention (ECA) module is introduced. Through extensive experiments, our model shows better performance on two publicly available radar datasets, CARRADA and RADDet, compared to other state-of-the-art methods. Compared with other mainstream methods that can only achieve 30–60% mAP performance when the IOU is 0.3, our model can achieve 74.51% and 75.62% on the RADDet and CARRADA datasets, respectively, and has better robustness and generalization ability.

show abstract

Review of Automatic Processing of Topography and Surface Feature Identification LiDAR Data Using Machine Learning Techniques

Cited by 28 publications

References 148 publications

A Review of Dynamic Object Filtering in SLAM Based on 3D LiDAR

A Review of Dynamic Object Filtering in SLAM Based on 3D LiDAR

Arbitrary-Oriented Object Detection in Aerial Images with Dynamic Deformable Convolution and Self-Normalizing Channel Attention

Radar Timing Range–Doppler Spectral Target Detection Based on Attention ConvLSTM in Traffic Scenes

Contact Info

Product

Resources

About