[1]郑 现,申 萍,邱 霁,等.视频拼接技术在列车安全监控中的应用[J].机车电传动,2018,(05):69-73.[doi:10.13890/j.issn.1000-128x.2018.05.016]
 ZHENG Xian,SHEN Ping,QIU Ji,et al.Video Stitching Technology in Railway Vehicle Safety Inspection[J].Electric Drive for Locomotives,2018,(05):69-73.[doi:10.13890/j.issn.1000-128x.2018.05.016]
点击复制

视频拼接技术在列车安全监控中的应用()
分享到:

机车电传动[ISSN:1000-128X/CN:43-1125/U]

卷:
期数:
2018年05期
页码:
69-73
栏目:
研究开发
出版日期:
2018-09-10

文章信息/Info

Title:
Video Stitching Technology in Railway Vehicle Safety Inspection
文章编号:
1000-128X(2018)05-0069-05
作者:
郑 现申 萍邱 霁夏顺盈
(北京交通大学 电气工程学院,北京 100044)
Author(s):
ZHENG Xian SHEN Ping QIU Ji XIA Shunying
( School of Electrical Engineering, Beijing Jiaotong University, Beijing 100044, China )
关键词:
视频拼接技术图像配准图像融合视频监控
Keywords:
video mosaic technology images matching images fusion video monitoring
分类号:
TN948.65;U298.1
DOI:
10.13890/j.issn.1000-128x.2018.05.016
文献标志码:
A
摘要:
传统列车监控系统中,采用多个独立摄像头对车厢内不同部位进行监控。这种监控画面存在视野狭窄、视角不一的问题,当出现突发事件时,需要在多个监控画面间来回切换,寻找和判断事故原因。为此提出一种基于拼接技术的车厢全景监控方案,采用的视频拼接核心算法中结合了PCA-SURF 和改进BBF 算法,并利用索引表法对多路视频帧序列进行实时拼接。试验结果表明,该算法相较于传统SURF 算法可将单帧拼接时间降低90%,正确匹配率平均达到92.2%,在实时性和准确性方面取得了良好的效果。
Abstract:
Traditional discrete monitoring system of train cabin used a number of independent cameras to monitor different parts inside. This monitoring method had narrow vision and different perspectives and staff members had to switch back and forth between multiple monitoring screens to find and judge the cause when there were sudden events.A panoramic monitoring scheme based on multi-camera video mosaic technology was proposed, which was the mosaic algorithm combining PCA-SURF and improved BBF. The index table was used to stitch video frame sequences from cameras in real time. The experimental results in laboratory environment showed that the proposed algorithm could reduce the time of single frame splicing by 90% compared with the traditional SURF algorithm,the correct matching rate reaching 92.2%, the proposed algorithm worked well in splicing speed and quality.

参考文献/References:

[1]SZELISKI R. Video mosaics for virtual environment[J]. IEEE Computer Graphics and Applications, 1996, 16(2): 22-30.
[2]LOWE D G. Distinctive image features from scale-invariant key points[J]. International Journal of Computer Vision, 2004, 60(2):91-110.
[3]BAY H, ESS A, TUYTELAARS T, et a1. Speeded-up robust features(SURF)[J]. Computer Vision and Image Understanding, 2008, 110(3):346-359.
[4] SHI X R, FAN Y S. Image mosaic algorithm for video surveillance [J].Electronic Test, 201l(12):8-11.
[5]李勇,杜丙新. 基于小区域融合的实时视频拼接技术[J]. 吉林大学学报,2016,54(6):1367-1372.
[6]PANCHAM A, WITHEY D, BRIGHT G. Tracking image features with PCA-SURF descriptors[C]//Iapr International Conference on Machine Vision Applications. Tokgo:IEEE, 2015:365-368.
[7]陈剑虹,韩小珍. 结合FAST-SURF 和改进k-d 树最近邻查找的图像配准[J]. 西安理工大学学报,2016,32(2):213-217.
[8]李丹,孙海涛,王海莉. 一种改进的SIFT 图像立体匹配算法[J]. 西南交通大学学报,2015,50(3):491-496.
[9]基于改进SURF 的实时视频拼接方法[J]. 计算机技术与发展, 2015,25(3):33-35.
 [10]JAIN P K, JAWAHAR CV. Homography estimation from planar contours[J]. Third International Symposium, 2006, 77(15): 877-884.
[11]Hartley, Richard. Multiple view geometry in computer vision[J]. Cambridge University Press, 2003,30(9/10):1865-1872.
[12]张敏. 基于多视域广角相机视频图像拼接技术研究[D]. 北京:中国科学院大学,2016.
[13]苗立刚. 图像与视频拼接算法研究[D]. 北京:中国科学院自动化研究所,2007.

备注/Memo

备注/Memo:
作者简介:郑 现(1993-),男,硕士研究生,研究方向为视频处理。
更新日期/Last Update: 2018-09-10