Topology-Aware Multi-View Street Scene Image Matching for Cross-Daylight Conditions Integrating Geometric Constraints and Semantic Consistency

Topology-Aware Multi-View Street Scene Image Matching for Cross-Daylight Conditions Integrating Geometric Constraints and Semantic Consistency

While deep learning-based image matching methods excel at extracting high-level semantic features from remote sensing data, their performance degrades significantly under cross-daylight conditions and wide-baseline geometric distortions, particularly in multi-source street-view scenarios. This paper...

Full description

Saved in:
Bibliographic Details
Main Authors: Haiqing He, Wenbo Xiong, Fuyang Zhou, Zile He, Tao Zhang, Zhiyuan Sheng
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:ISPRS International Journal of Geo-Information
Subjects:
image matching
deep learning
geometric constraint
semantic consistency
cross-daylight conditions
Online Access:https://www.mdpi.com/2220-9964/14/6/212
Tags: Add Tag
No Tags, Be the first to tag this record!

Internet

https://www.mdpi.com/2220-9964/14/6/212

Similar Items