We present and discuss the case for full waveform pixel and image acquisition and processing to enable LiDAR sensors to penetrate and reconstruct 3D surface maps through obscuring media. To that end, we review work on signal propagation, on scanning and arrayed sensors, on signal processing strategies for independent pixels and employing spatial context, on reducing complexity and accelerating processing by sensor design, algorithmic changes, compressed sensing, and parallel processing. We report several experimental studies on LiDAR imaging through complex media, and how these can inform the automotive LiDAR scenario. We conclude with a discussion of future development and potential for full waveform LiDAR (FWL).

/pdf/full-waveform-lidar-for-adverse-weather-conditions-2e35q3mkh0.pdf

Full Waveform LiDAR for Adverse Weather Conditions

Light detection and ranging sensors (LiDARS) are the most promising devices for range sensing in automated cars and therefore, have been under intensive development for the last five years. Even though various types of resolutions and scanning principles have been proposed, adverse weather conditions are still challenging for optical sensing principles. This paper investigates proposed methods in the literature and adopts a common validation method to perform both indoor and outdoor tests to examine how fog and snow affect performances of different LiDARs. As suspected, the performance degraded with all tested sensors, but their behavior was not identical.

/pdf/testing-and-validation-of-automotive-point-cloud-sensors-in-1gj1msotw0.pdf

Testing and validation of automotive point-cloud sensors in adverse weather conditions

Multi-object detection and multi-object-tracking in diverse driving situations is the main challenge in autonomous vehicles. Vehicle manufacturers and research organizations are addressing this problem, with multiple sensors such as camera, LiDAR, RADAR, ultrasonic-sensors, GPS, and Vehicle-to-Everything-technology. Deep Neural Networks (DNN) are playing a predominant role to solve this. Fusing the sensing modalities with DNN will be the leading solution to this challenge. This paper evaluates the state-of-the-art techniques that address this challenge, with three primary sensors camera, LiDAR, and RADAR with DNN, and fusion of sensor data with DNN. The analysis shows that there exists an excellent potential to design a more optimized solution to address this challenge. This work proposes a perception model for autonomous vehicles.

Multi-Object Detection and Tracking, Based on DNN, for Autonomous Vehicles: A Review

Datasets for autonomous cars are essential for the development and benchmarking of perception systems. However, most existing datasets are captured with camera and LiDAR sensors in good weather conditions. In this paper, we present the RAdar Dataset In Adverse weaThEr (RADIATE), aiming to facilitate research on object detection, tracking and scene understanding using radar sensing for safe autonomous driving. RADIATE includes 3 hours of annotated radar images with more than 200K labelled road actors in total, on average about 4.6 instances per radar image. It covers 8 different categories of actors in a variety of weather conditions (e.g., sun, night, rain, fog and snow) and driving scenarios (e.g., parked, urban, motorway and suburban), representing different levels of challenge. To the best of our knowledge, this is the first public radar dataset which provides high-resolution radar images on public roads with a large amount of road actors labelled. The data collected in adverse weather, e.g., fog and snowfall, is unique. Some baseline results of radar based object detection and recognition are given to show that the use of radar data is promising for automotive applications in bad weather, where vision and LiDAR can fail. RADIATE also has stereo images, 32-channel LiDAR and GPS data, directed at other applications such as sensor fusion, localisation and mapping. The public dataset can be accessed at this http URL

RADIATE: A Radar Dataset for Automotive Perception in Bad Weather

Low-terahertz (Low-THz, 100 GHz-1.0 THz) technology is expected to provide unprecedented data rates in future generations of wireless system such as the 6th generation (6G) mobile communication system. Increasing the carrier frequencies from millimeter wave to THz is a potential solution to guarantee the transmission rate and channel capacity. Due to the large transmission loss of Low-THz wave in free space, it is particularly urgent to design high-gain antennas to compensate the additional path loss, and to overcome the power limitation of Low-THz source. Recently, with the continuous updating and progress of additive manufacturing (AM) and 3D printing (3DP) technology, antennas with complicated structures can now be easily manufactured with high precision and low cost. In the first part, this paper demonstrates different approaches of recent development on wideband and high gain sub-millimeter-wave and Low-THz antennas as well as their fabrication technologies. In addition, the performances of the state-of-the-art wideband and high-gain antennas are presented. A comparison among these reported antennas is summarized and discussed. In the second part, one case study of a broadband high-gain antenna at 300 GHz is introduced, which is an all-metal model based on the Fabry–Perot cavity (FPC) theory. The proposed FPC antenna is very suitable for manufacturing using AM technology, which provides a low-cost, reliable solution for emerging THz applications.

/pdf/a-review-of-broadband-low-cost-and-high-gain-low-terahertz-3kp6y104te.pdf

A Review of Broadband Low-Cost and High-Gain Low-Terahertz Antennas for Wireless Communications Applications

Low-THz Radar, Lidar and Optical Imaging through Artificially Generated Fog

In this study, the authors investigate the application of the Doppler beam sharpening (DBS) technique for angular refinement to the emerging area of low-terahertz (THz) radar sensing. Ultimately this is to improve radar image quality in the azimuth plane to complement the excellent range resolution and thus improve object classification in low-THz radar imaging systems for autonomous platforms. The study explains the fundamental theory behind the process of DBS and describes the applicability of DBS to automotive sensing, indicating the potential for synthetic beamwidths of a fraction of a degree. Low-THz DBS was experimentally tested under controlled laboratory conditions, not only to accurately localised target objects in Cartesian space but also to provide unique object imaging at low-THz frequencies with wide azimuthal beamwidth antennas. It was shown that a stationary (i.e. non-scanned) wide beam antenna mounted on a moving platform can deliver imagery at least comparable to that produced by physical beamforming, be that steering arrays or narrow beam scanning antennas as in the experimental case presented.

Application of Doppler Beam Sharpening for Azimuth Refinement in Prospective Low-THz Automotive Radars

This paper presents a solution to the current challenges of the imaging radar to respond the demands of autonomy for detection and classification of targets in radar imagery, which traditionally has been considered as clutter. The proposed object detection method is defined in a new way, as opposed to the traditional object detection methods in the radar related contexts. The current paper presents the first application of this novel approach, based on deep neural networks for object detection, on outdoor radar images, as well as indoor images taken in controlled environment. Object detection was performed using two detectors, Faster R-CNN and SSD and the evaluation proved that this method can be successfully used on radar imagery for autonomous applications.

Object Detection on Radar Imagery for Autonomous Driving Using Deep Neural Networks

In this paper we present a series of road scene images recorded with a prototype high resolution low-THz imaging radar. The aim of this study is to assess the images in order to identify and extract features of interest for scene characterisation to aid the development of future autonomous vehicles. Images were recorded at multiple radar heights in order to assess the effect of illumination angle on the selected features. At this stage of research the recorded images are presented and more qualitative analysis of a subset of the features is given.

Dominic Phippen

Papers

Low-THz Radar, Lidar and Optical Imaging through Artificially Generated Fog

Application of Doppler Beam Sharpening for Azimuth Refinement in Prospective Low-THz Automotive Radars

Object Detection on Radar Imagery for Autonomous Driving Using Deep Neural Networks

Multi-height radar images of road scenes at 150 GHz

Trilateration of Targets Using a 300GHz Radar System