A New Object Detection Method in Indoor Scenes Based on Spatial Distance Clustering

DOI

10.2991/cnci-19.2019.90How to use a DOI?

Keywords

Deep learning, point cloud, object detection.

Abstract

3D object detection is one of the important problems in machine vision. With the increasing popularity of depth camera, 3d object detection in point cloud has become a research hotspot of 3D vision. The irregular format of point cloud makes the traditional deep learning method based on image convolution unable to understand and analyze it. In recent years, the deep learning framework as PointNet[1] directly applied to the original structure of point cloud has greatly improved the ability of deep learning network to process point cloud data. This paper depends on the existing point cloud deep learning framework, proposes an object detection method based on semantic segmentation. We use the clustering method based on region growth to detect the objects in indoor scenes. This method has been tested on the Stanford large-scale 3D Indoor Spaces Dataset (S3DIS) Dataset[2] with good results.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Dianyuan Wu
AU  - Kai Xu
PY  - 2019/05
DA  - 2019/05
TI  - A New Object Detection Method in Indoor Scenes Based on Spatial Distance Clustering
BT  - Proceedings of the 2019 International Conference on Computer, Network, Communication and Information Systems (CNCI 2019)
PB  - Atlantis Press
SP  - 650
EP  - 654
SN  - 2352-538X
UR  - https://doi.org/10.2991/cnci-19.2019.90
DO  - 10.2991/cnci-19.2019.90
ID  - Wu2019/05
ER  -