Segmentation of Lidar Intensity using Weighted Fusion based on Appropriate Region Size

Masaki Umemura, Kazuhiro Hotta, Hideki Nonaka, Kazuo Oda

2018

Abstract

We propose a semantic segmentation method for LiDAR intensity images obtained by Mobile Mapping System (MMS). Conventional segmentation method could give high pixel-wise accuracy but the accuracy of small objects is quite low. We solve this issue by using the weighted fusion of multi-scale inputs because each class has the most effective scale that small object class gives higher accuracy for small input size than large input size. In experiments, we use 36 LIDAR intensity images with ground truth labels. We divide 36 images into 28 training images and 8 test images. Our proposed method gain 87.41% on class average accuracy, and it is 5% higher than conventional method. We demonstrated that the weighted fusion of multi-scale inputs is effective to improve the segmentation accuracy of small objects.

Download


Paper Citation


in Harvard Style

Umemura M., Hotta K., Nonaka H. and Oda K. (2018). Segmentation of Lidar Intensity using Weighted Fusion based on Appropriate Region Size.In Proceedings of the 7th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-276-9, pages 608-613. DOI: 10.5220/0006717706080613


in Bibtex Style

@conference{icpram18,
author={Masaki Umemura and Kazuhiro Hotta and Hideki Nonaka and Kazuo Oda},
title={Segmentation of Lidar Intensity using Weighted Fusion based on Appropriate Region Size},
booktitle={Proceedings of the 7th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2018},
pages={608-613},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006717706080613},
isbn={978-989-758-276-9},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 7th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Segmentation of Lidar Intensity using Weighted Fusion based on Appropriate Region Size
SN - 978-989-758-276-9
AU - Umemura M.
AU - Hotta K.
AU - Nonaka H.
AU - Oda K.
PY - 2018
SP - 608
EP - 613
DO - 10.5220/0006717706080613