Skip to main content

2024 | Book

Robust Environmental Perception and Reliability Control for Intelligent Vehicles

Authors: Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao

Publisher: Springer Nature Singapore

Book Series : Recent Advancements in Connected Autonomous Vehicle Technologies


About this book

This book presents the most recent state-of-the-art algorithms on robust environmental perception and reliability control for intelligent vehicle systems. By integrating object detection, semantic segmentation, trajectory prediction, multi-object tracking, multi-sensor fusion, and reliability control in a systematic way, this book is aimed at guaranteeing that intelligent vehicles can run safely in complex road traffic scenes.Adopts the multi-sensor data fusion-based neural networks to environmental perception fault tolerance algorithms, solving the problem of perception reliability when some sensors fail by using data redundancy.Presents the camera-based monocular approach to implement the robust perception tasks, which introduces sequential feature association and depth hint augmentation, and introduces seven adaptive methods.Proposes efficient and robust semantic segmentation of traffic scenes through real-time deep dual-resolution networks and representation separation of vision transformers.Focuses on trajectory prediction and proposes phased and progressive trajectory prediction methods that is more consistent with human psychological characteristics, which is able to take both social interactions and personal intentions into account.Puts forward methods based on conditional random field and multi-task segmentation learning to solve the robust multi-object tracking problem for environment perception in autonomous vehicle scenarios. Presents the novel reliability control strategies of intelligent vehicles to optimize the dynamic tracking performance and investigates the completely unknown autonomous vehicle tracking issues with actuator faults.

Table of Contents

Chapter 1. Background
The key aspects of intelligent vehicle systems are increased performance and safety in complex road traffic scenarios, which are achieved through the use of environmental perception and control systems. Intelligent vehicles that possess these capabilities rely heavily on robust environmental perception and reliable control systems. To improve the reliability and availability of intelligent vehicles while maintaining satisfactory performance, it is essential to design robust perception and reliability control systems that can guarantee safe operation. In this chapter, we introduce the background knowledge to design such systems, which illustrate the reliability capabilities of intelligent vehicles through robust perception and control schemes.
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Chapter 2. Robust Environmental Perception of Multi-sensor Data Fusion
In this chapter, the multi-sensor data fusion based neural networks are adopted to environmental perception fault tolerance algorithms. Multi-sensor fusion can solve the problem of perception reliability when some sensors fail by using data redundancy and improve the robustness of the perception algorithms. In Sect. 2.1, a 2D-3D pose estimation network based on keypoints is proposed, which can be applied to calibrate the camera and LiDAR in real-time. In Sect. 2.2, a real-time data fusion network with fault diagnosis and fault tolerance mechanisms is designed. By leveraging temporal and spatial correlations between sensor data, this network utilizes sensor redundancy to diagnose local and global confidence of sensor data in real-time, eliminating faulty data and ensuring accuracy and reliability of data fusion. In Sect. 2.3, a novel multi-phase fusion network for robust 3D semantic segmentation is proposed. Three factors that restrict the performance of fusion-based 3D semantic segmentation methods are summarized: inefficient feature fusion mechanism, inability to effectively express features, and lacking of dense labels, and corresponding solutions are put forward. The usefulness and advantage of the proposed algorithms are demonstrated via experiments on common datasets and real scenes.
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Chapter 3. Robust Environmental Perception of Monocular 3D Object Detection
The main role of monocular vision based 3D object detection in intelligent vehicles is to extract obstacle information from the environment, including the category of object, location and orientation. Utilizing monocular cameras for this purpose has several advantages, such as being cost-effective and easily accessible compared to other types of sensors. The realization of 3D object detection through monocular vision alone has considerable potential for both commercial and research purposes. In the context of intelligent vehicles, perception systems play a key role by accurately assessing the surrounding environmental conditions. These systems generate reliable observations, which are essential for tasks such as prediction and planning. Within this framework, 3D object detection is a key feature that allows systems to predict the location, size and class of important 3D objects in the vicinity of intelligent vehicles. This chapter of the book is dedicated to this task. In order to address the different aspects of accuracy and robustness, two approaches, FANet and AMNet, are introduced.
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Chapter 4. Robust Environmental Perception of Semantic Segmentation
Semantic segmentation is a fundamental computer vision task and has a wide range of applications [13].
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Chapter 5. Robust Environmental Perception of Trajectory Prediction
Trajectory prediction has always been an important research direction in the field of transportation (Helbing and Molnar in Phys Rev E 51(5):4282, 1995, [1]; Luber et al. in People tracking with human motion predictions from social forces. IEEE, pp 464–469, 2010, [2]; Mehran et al. in Abnormal crowd behavior detection using social force model. IEEE, pp 935–942, 2009, [3]).
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Chapter 6. Robust Environmental Perception of Multi-object Tracking
Multi-object tracking (MOT) is an essential component in a number of fields such as robotics, autonomous vehicles, and surveillance systems. For accurate tracking and predicting of the motion of multiple objects in complex environments, strong sensing capabilities are crucial.
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Chapter 7. Reliability Control of Intelligent Vehicles
The reliability of control strategies becomes more important in the case of intelligent Vehicles. The traditional vehicle controllers are usually based on the assumption that the parameters of the vehicle system are fixed constants or the dynamic is linear systems. These approaches can simplify the design steps of the controller and make the control structure simple. However, the actual vehicle system are nonlinear, and the vehicle system parameters are also unknown and variable.
Huihui Pan, Jue Wang, Xinghu Yu, Weichao Sun, Huijun Gao
Robust Environmental Perception and Reliability Control for Intelligent Vehicles
Huihui Pan
Jue Wang
Xinghu Yu
Weichao Sun
Huijun Gao
Copyright Year
Springer Nature Singapore
Electronic ISBN
Print ISBN

Premium Partner