Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ERASOR++: Height Coding Plus Egocentric Ratio Based Dynamic Object Removal for Static Point Cloud Mapping

Jiabao Zhang, Yu Zhang

TL;DR

This work proposes ERASOR++, an enhanced approach based on the Egocentric Ratio of Pseudo Occupancy for effective dynamic object removal and introduces the Height Coding Descriptor, which combines height difference and height layer information to encode the point cloud.

Abstract

Mapping plays a crucial role in location and navigation within automatic systems. However, the presence of dynamic objects in 3D point cloud maps generated from scan sensors can introduce map distortion and long traces, thereby posing challenges for accurate mapping and navigation. To address this issue, we propose ERASOR++, an enhanced approach based on the Egocentric Ratio of Pseudo Occupancy for effective dynamic object removal. To begin, we introduce the Height Coding Descriptor, which combines height difference and height layer information to encode the point cloud. Subsequently, we propose the Height Stack Test, Ground Layer Test, and Surrounding Point Test methods to precisely and efficiently identify the dynamic bins within point cloud bins, thus overcoming the limitations of prior approaches. Through extensive evaluation on open-source datasets, our approach demonstrates superior performance in terms of precision and efficiency compared to existing methods. Furthermore, the techniques described in our work hold promise for addressing various challenging tasks or aspects through subsequent migration.

ERASOR++: Height Coding Plus Egocentric Ratio Based Dynamic Object Removal for Static Point Cloud Mapping

TL;DR

This work proposes ERASOR++, an enhanced approach based on the Egocentric Ratio of Pseudo Occupancy for effective dynamic object removal and introduces the Height Coding Descriptor, which combines height difference and height layer information to encode the point cloud.

Abstract

Mapping plays a crucial role in location and navigation within automatic systems. However, the presence of dynamic objects in 3D point cloud maps generated from scan sensors can introduce map distortion and long traces, thereby posing challenges for accurate mapping and navigation. To address this issue, we propose ERASOR++, an enhanced approach based on the Egocentric Ratio of Pseudo Occupancy for effective dynamic object removal. To begin, we introduce the Height Coding Descriptor, which combines height difference and height layer information to encode the point cloud. Subsequently, we propose the Height Stack Test, Ground Layer Test, and Surrounding Point Test methods to precisely and efficiently identify the dynamic bins within point cloud bins, thus overcoming the limitations of prior approaches. Through extensive evaluation on open-source datasets, our approach demonstrates superior performance in terms of precision and efficiency compared to existing methods. Furthermore, the techniques described in our work hold promise for addressing various challenging tasks or aspects through subsequent migration.
Paper Structure (16 sections, 8 equations, 6 figures, 2 tables)

This paper contains 16 sections, 8 equations, 6 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Dynamic Object Removal Methods
  4. Main Works and Limitations in ERASOR
  5. LiDAR Descriptors
  6. Methodology in ERASOR++
  7. Height Coding Descriptor
  8. Height Stack Test
  9. Ground Layer Test
  10. Surrounding Points Test
  11. Experiments and Discussion
  12. Settings and Datasets
  13. Algorithm Precision
  14. Algorithm Speed
  15. Ablation Experiments
  16. ...and 1 more sections

Figures (6)

  • Figure 1: Framework of ERASOR++. The main parts are labeled with corresponding sections, with respective resolved limitations mentioned in related works. Note that the processes before the VoI part and after the R-GPF part transfer from the previous work ERASORerasor.
  • Figure 2: The representation of VoI range and Bin division of point cloud, along with the definition of combined Height Difference Descriptor and Height Encoding Descriptor, as well as a numeral example to clarify encoding part.
  • Figure 3: The good removal and bad removal, which can not be distinguished through Scan Ratio Test, but can be greatly separated through Height Stack Test, according to the formula and example shown in this figure. Note that $threshold$ is a constant value and empirically set to 0.2 in erasor. When the overlap layer is only around the ground, the dynamic removal is good and valid. On the contrary, when the overlap is in high layers, there may be bad removal.
  • Figure 4: (Left) Isolated dynamic bin in point cloud, which may cause bad removal. (Right) The searching area of each dynamic bin.
  • Figure 5: Visual Comparison of Experiments Results using ERASOR and ERASOR++. Through the main difference in red boxes, our proposed method can largely preserve static structures, such as buildings and vegetation.
  • ...and 1 more figures