Abstract
With the rise of advanced driver assistance systems (ADAS), range sensors and their data processing methods are becoming more and more important. Light detection and ranging (LiDAR) sensors are attracting attention due to their unique advantages in terms of radial distance resolution and detection range. However, the study of LiDAR data processing is usually divorced from the LiDAR sensor measurement process itself. This leads to critical measurement information being overlooked. This paper seeks a breakthrough to improve the performance of single-photon-avalanche-diode-based direct time-of-flight LiDAR systems by reviewing the data processing stages and corresponding processing approaches for LiDAR measurements, starting from photon incidence and ending with high-level feature recognition. Firstly, we propose a LiDAR system model based on data generation and transfer. The data forms in such a LiDAR system are mainly classified into timestamps, time-correlated histograms, point cloud data, and high-level properties. Subsequently, data processing methods applied to each of these data forms are analyzed. A number of hardware solutions closely related to data transmission and control are also included in the discussion. The principles, limitations, and challenges of these methods are discussed in detail and the criteria for evaluation of time-correlated histograms in ADAS are proposed. Finally, the research gaps in data processing are summarized, and future directions for research development are presented.
Highlights
T RAFFIC safety has always been a topic of great concern
In the flash Light detection and ranging (LiDAR) system, one frame is equal to the completion time for one TC-HIST
Considering that one measurement period is extremely short and that the detection array contains at least hundreds of SPADS, a LiDAR system can generate large amounts of data each second
Summary
T RAFFIC safety has always been a topic of great concern. In 2016, it is reported that 1,099,032 traffic accidents took place, resulting in 25,651 fatalities and more than 1.4 million injured people in Europe [1]. The main contributions of this work are: 1) to provide an overview of the current state of development by sorting out existing approaches on the LiDAR data processing workflow, 2) to point out weaknesses or areas that have not been covered by research, 3) to discuss future research directions and advance the development of data processing. Towards this end, the review is structured as follows: in section II, a data-transfer-based LiDAR system model is proposed.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
