Adaptive and Efficient Mixture-Based Representation for Range Data.

Minghe Cao,Jianzhong Wang,Li Ming

doi:10.3390/s20113272

Abstract

Modern range sensors generate millions of data points per second, making it difficult to utilize all incoming data effectively in real time for devices with limited computational resources. The Gaussian mixture model (GMM) is a convenient and essential tool commonly used in many research domains. In this paper, an environment representation approach based on the hierarchical GMM structure is proposed, which can be utilized to model environments with weighted Gaussians. The hierarchical structure accelerates training by recursively segmenting local environments into smaller clusters. By adopting the information-theoretic distance and shape of probabilistic distributions, weighted Gaussians can be dynamically allocated to local environments in an arbitrary scale, leading to a full adaptivity in the number of Gaussians. Evaluations are carried out in terms of time efficiency, reconstruction, and fidelity using datasets collected from different sensors. The results demonstrate that the proposed approach is superior with respect to time efficiency while maintaining the high fidelity as compared to other state-of-the-art approaches.

Highlights

Range data have widely been used in applications including medical imaging, object modeling, and robotics state estimation
The Corner and Office are from the TUM dataset [41], whose range data are collected by a Microsoft Kinect RGB-D camera
To assess the grid map constructed by the resampled points, we take the map created by raw point cloud as ground truth, query the status of the two maps with a set of uniformly distributed points

Summary

Introduction

Range data have widely been used in applications including medical imaging, object modeling, and robotics state estimation. Modern range sensors (e.g., LiDAR, RGB-D cameras) generate millions of data points per second, making it difficult to utilize all incoming data effectively in real time for devices with limited computational resources. Along with the impact of sensor noises, it is rarely feasible to operate directly on the raw point measurements obtained from the sensors in robotics applications. Point cloud representation techniques are developed to address this problem. The direct dense representation involves stitching point clouds from different poses of frames, and downsampling data via various filters. Other techniques include discretizing the environment into grids or voxels

Methods

Results

Discussion

Conclusion