首页 | 本学科首页   官方微博 | 高级检索  
     

基于准不动点集的高速铁路前向运动视频区域划分方法
引用本文:蒋欣兰,尹辉,罗四维,景璞,吕国豪. 基于准不动点集的高速铁路前向运动视频区域划分方法[J]. 北京交通大学学报(自然科学版), 2017, 41(2). DOI: 10.11860/j.issn.1673-0291.2017.02.006
作者姓名:蒋欣兰  尹辉  罗四维  景璞  吕国豪
作者单位:北京交通大学交通数据分析与挖掘北京市重点实验室,北京,100044;济南铁路局,济南,250001
基金项目:国家自然科学基金,北京市自然科学基金(4152042)National Natural Science Foundation of China,Beijing Municipal Natural Science Foundation
摘    要:为保障高速铁路行车安全,以快速检测高速铁路沿线运行环境为背景,建立了基于高速前向运动视频的环境检测模型,并提出人工干预的前向运动视频全景图拼接方法.该模型包括高速前向运动视频区域划分、全景图拼接等.为实现前向运动视频区域自动区域划分,本文提出一种基于准不动点集的前向运动视频消失点定位方法,结合铁路线路的先验信息可实现前向运动视频区域自动划分.实验结果表明:该方法是一种无需人工干预的快速自动划分方法,为铁路前向运动视频全景图全自动拼接提供参考.

关 键 词:准不动点集  区域划分  前向运动视频

Region partition method for high-speed railway forward-looking motion video based on quasi-fixed point set
JIANG Xinlan,YIN Hui,LUO Siwei,JING Pu,LYU Guohao. Region partition method for high-speed railway forward-looking motion video based on quasi-fixed point set[J]. JOURNAL OF BEIJING JIAOTONG UNIVERSITY, 2017, 41(2). DOI: 10.11860/j.issn.1673-0291.2017.02.006
Authors:JIANG Xinlan  YIN Hui  LUO Siwei  JING Pu  LYU Guohao
Abstract:In order to ensure the safety of high-speed railway,a new railway running environment inspection model based on high-speed forward-looking motion video has been established,and a panorama stitching approach has been proposed with human intervention.The model includes the region partition and panoramic stitching.In order to achieve automatic partition in the high-speed forward-looking motion video,this paper proposes a novel method for locating vanishing point of the forward-looking motion video based on quasi-fixed point set,which can be combined with the priori information of the railway line,we can realize automatic partition of the forward-looking motion video.The experimental results show that this method is a fast automatic method without manual intervention,which lays a foundation for automatic panorama stitching of railway forward-looking motion video.
Keywords:quasi-fixed point set  region partition  forward-looking motion video
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号