On-Road Object Detection and Tracking Based on Radar and Vision Fusion: A Review

Zhang, Zhiqiang; Qin, Yechen

doi:10.1109/mits.2021.3093379

Cited by 54 publications

(18 citation statements)

References 159 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, they do not cover the entire near-field, and there are some blind spots. Radar is additionally limited in that it cannot detect road markings and has limited performance in object classification [26]. Parking space detection using SRR is discussed in more detail in [27].…”

Section: B Relation To Other Sensorsmentioning

confidence: 99%

“…Varga et al [30] have attempted to combine fisheye camera and LiDAR to provide a unified 360 • environmental model, but there are blind spots in the near field. Classification of objects in LiDAR has extremely limited performance [26]. To summarize, other near-field sensors like radar and sonar capture limited information about the scene, and thus they cannot operate independently to perform near-field perception.…”

Section: B Relation To Other Sensorsmentioning

confidence: 99%

See 1 more Smart Citation

Surround-View Fisheye Camera Perception for Automated Driving: Overview, Survey & Challenges

Kumar

Eising

Witt

et al. 2023

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

Surround-view fisheye cameras are commonly used for near-field sensing in automated driving. Four fisheye cameras on four sides of the vehicle are sufficient to cover 360 • around the vehicle capturing the entire near-field region. Some primary use cases are automated parking, traffic jam assist, and urban driving. There are limited datasets and very little work on near-field perception tasks as the focus in automotive perception is on far-field perception. In contrast to far-field, surround-view perception poses additional challenges due to high precision object detection requirements of 10cm and partial visibility of objects. Due to the large radial distortion of fisheye cameras, standard algorithms cannot be extended easily to the surround-view use case. Thus, we are motivated to provide a self-contained reference for automotive fisheye camera perception for researchers and practitioners. Firstly, we provide a unified and taxonomic treatment of commonly used fisheye camera models. Secondly, we discuss various perception tasks and existing literature. Finally, we discuss the challenges and future direction.

show abstract

Section: B Relation To Other Sensorsmentioning

confidence: 99%

Section: B Relation To Other Sensorsmentioning

confidence: 99%

Surround-View Fisheye Camera Perception for Automated Driving: Overview, Survey & Challenges

Kumar

Eising

Witt

et al. 2023

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

show abstract

“…As shown in Figure 3, the input of the integrated framework is the COM trajectory motion reference, which can be generated from the road terrains based on radar and vision fusion. 33 In this paper, all four wheels are maintained in contact with the ground under traveling on different terrains. To clearly describe the parameters of the attitude control, we define the other corner notations for the relevant parameters, where the right-hand corner cmd denotes the control command parameters and the right-hand corner fb and ff denote the feedback parameter and feedforward parameter.…”

Section: Attitude Controller Designmentioning

confidence: 99%

Attitude control strategy for unmanned wheel-legged hybrid vehicles considering the contact of the wheels and ground

Liu

Han

et al. 2021

Proceedings of the Institution of Mechanical Engineers, Part D:

Self Cite

View full text Add to dashboard Cite

During patrol and surveillance tasks, attitude control is crucial for improving the terrain adaptability of unmanned wheel-legged hybrid vehicles. This paper proposes an attitude control strategy for unmanned wheel-legged hybrid vehicles, considering the contact of the wheels and ground. The proposed method can naturally achieve torque control efficiently of each joint actuator and wheel-side actuator and avoid the discrepancy between off-road terrain and stability. First, an inverse kinematics model is established to resolve the body and each joint rotation angle, and the dynamic model is built based on the multi rigid body theory, considering the contact points planning of wheel and ground. Considering the nonholonomic constraint of the structure scheme, a hierarchical real-time attitude controller for a wheel-legged vehicle is proposed. The upper layer calculates the contact points of each wheel and the ground through the quadratic programming algorithm, and the lower layer is divided into a legged motion generator and a wheel motion generator by a mathematical analysis method. Finally, the proposed method is applied to achieve the tracking and control of the whole-body trajectory. The proposed strategy can achieve the decoupling of wheeled motion generator and legged motion generator, and improve control efficiency.

show abstract

“…Besides the aforementioned applications, radars have also been used for vehicle position estimation [10]- [14], and detection of other vehicles aided by Light Detection And Ranging (LiDAR) sensors [15] or cameras [16]. Usually, combining radar scans with other sensors is interesting given that their operating conditions are complementary [17].…”

Section: Introductionmentioning

confidence: 99%

Towards Land Vehicle Forward Velocity Estimation using Deep Learning and Onboard Radars

Araujo¹,

Noureldin²,

Givigi³

2023

Preprint

View full text Add to dashboard Cite

<p>Radars have been increasingly implemented in modern vehicles. Their presence opens opportunities for developing new positioning and collision avoidance solutions, to cite a few. However, radar scans are noisy and sparse, which challenges the robustness of the developed solutions. In this paper, we propose a novel method to structure radar scans to create point descriptors. The structured data is used in convolutional neural networks that explore orthogonal views of the radar point cloud. Experimental results demonstrate that the model performs well in estimating the forward velocity of the vehicle using only the radar scans, providing estimations at a higher data rate than odometers available in the vehicles.</p>

show abstract

On-Road Object Detection and Tracking Based on Radar and Vision Fusion: A Review

Cited by 54 publications

References 159 publications

Surround-View Fisheye Camera Perception for Automated Driving: Overview, Survey & Challenges

Surround-View Fisheye Camera Perception for Automated Driving: Overview, Survey & Challenges

Attitude control strategy for unmanned wheel-legged hybrid vehicles considering the contact of the wheels and ground

Towards Land Vehicle Forward Velocity Estimation using Deep Learning and Onboard Radars

Contact Info

Product

Resources

About