Monocular-Based Vehicle Displacement Detection in Low-Light Dynamic Environments

View at publisher Summarise with VIRAT
Authors: Kezhong Liu, Cong Fan, Shengkai Zhang, C.X. Li, Xuedou Xiao, Xuming Zeng, Mozi Chen
Cited by: 0
References: 14

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

In this letter, we propose a monocular–based vehicle displacement detection system for low-light dynamic environments in transport vessels. The proposed system integrates two key designs: (1) an image enhancement module with resolution-adaptive feature fusion and uncertainty-guided correction to improve sensing reliability under harsh lighting, and (2) a motion compensation module that combines global alignment with confidence-weighted optical-flow refinement to suppress jitter-induced pseudo-motion and capture fine-scale vehicle displacement. Extensive experiments on real world transport-ship datasets collected at Lianyungang Port, Jiangsu Province, China, demonstrate that our system achieves a displacement error of 4 cm and reduces error by 60.2% compared to representative state-of-the-art baselines.

Fields of Study & Subjects

Video Surveillance and Tracking Methods Image and Video Stabilization Advanced Vision and Imaging

Publication details

Journal / source
IEEE Sensors Letters
Volume
10
Issue
5
Pages
1–4
Publication date
23 March 2026
Language
EN