Selecting Optimal Combination of Data Channels for Semantic Segmentation in City Information Modelling (CIM)

Cai, Yuanzhi ORCID: 0000-0002-7005-5870, Huang, Hong, Wang, Kaiyang ORCID: 0000-0003-2420-5521, Zhang, Cheng ORCID: 0000-0003-2465-1767, Fan, Lei and Guo, Fangyu (2021) Selecting Optimal Combination of Data Channels for Semantic Segmentation in City Information Modelling (CIM). REMOTE SENSING, 13 (7). p. 1367.

Text
Selecting Optimal Combination of Data Channels for Semantic Segmentation in City Information Modelling (CIM).pdf - Open Access published version
Download (20MB) | Preview

Official URL: http://dx.doi.org/10.3390/rs13071367

Abstract

<jats:p>Over the last decade, a 3D reconstruction technique has been developed to present the latest as-is information for various objects and build the city information models. Meanwhile, deep learning based approaches are employed to add semantic information to the models. Studies have proved that the accuracy of the model could be improved by combining multiple data channels (e.g., XYZ, Intensity, D, and RGB). Nevertheless, the redundant data channels in large-scale datasets may cause high computation cost and time during data processing. Few researchers have addressed the question of which combination of channels is optimal in terms of overall accuracy (OA) and mean intersection over union (mIoU). Therefore, a framework is proposed to explore an efficient data fusion approach for semantic segmentation by selecting an optimal combination of data channels. In the framework, a total of 13 channel combinations are investigated to pre-process data and the encoder-to-decoder structure is utilized for network permutations. A case study is carried out to investigate the efficiency of the proposed approach by adopting a city-level benchmark dataset and applying nine networks. It is found that the combination of IRGB channels provide the best OA performance, while IRGBD channels provide the best mIoU performance.</jats:p>

Item Type:	Article
Uncontrolled Keywords:	3D reconstruction, city information modelling, data channels, data fusion, point cloud, semantic segmentation
Divisions:	Faculty of Science and Engineering > School of Engineering
Depositing User:	Symplectic Admin
Date Deposited:	13 Feb 2024 11:31
Last Modified:	16 Mar 2024 00:11
DOI:	10.3390/rs13071367
Related URLs:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3178631