Overview of object detection using multi-source images for railway perimeter protection
YU Zujun
ZHOU Xingfang
GUO Baoqing
BAI Dingyuan
Abstract：Railway perimeter protection is crucial for ensuring the safety of railway transportation,and video surveillance stands out as the most widely utilized technology in this domain.Aiming to address the issue of high false positive and false negative rates under challenging lighting conditions,particu-larly at night,given the ability of infrared imaging to capture thermal radiation from objects and its re-silience against variations in lighting and adverse weather,this study investigates the integration of in-frared and visible light images to enable all-weather perimeter intrusion detection.Firstly,based on an analysis of the current state of railway perimeter protection technology,the paper discusses the focal points and challenges in research related to object detection through the fusion of infrared and visible light images.Subsequently,it outlines the general methods and procedures for image registration and object detection,while analyzing the progress made in infrared image registration,infrared target de-tection,and the research advancements in fusion-based object detection using infrared and visible light images in railway scenarios.Finally,this paper provides an outlook on the future trends in multi-source image object detection technology within the context of railway perimeter protection applica-tions.The combination of infrared and visible light video images not only facilitates effective intrusion detection under all-weather conditions but also ensures the visual representation of intrusion targets,thereby facilitating subsequent intelligent analysis.
Keywords：railway transportationperimeter protectionheterogeneous image registrationobject de-tection based on multi-source image
Publication Date：2023-10-28
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：15( 1-15 )