基于FAST特征检测和FLANN特征点匹配算法的隧道展布图拼接方法研究  

Tunnel Layout Map Stitching Method Research Based on FAST Feature Detectionand FLANN Feature Point Matching Algorithm

作　　者：廖勇 黄伟宏 钟芸 LIAO Yong;HUANG Weihong;ZHONG Yun(China Railway Changjiang Transport Design Group Co.,Ltd,Chongqing 400067;Chongqing Transportation Engineering Quality Inspection Co.,Ltd.,Chongqing 400074,China)

机构地区：[1]中铁长江交通设计集团有限公司,重庆400067 [2]重庆市交通工程质量检测有限责任公司,重庆400074

出　　处：《交通节能与环保》2021年第2期137-141,共5页Transport Energy Conservation & Environmental Protection

基　　金：重庆市技术创新与应用示范专项项目(cstc2018jszx-cyzdX0134)。

摘　　要：本文研究基于FAST特征检测和FLANN特征点匹配算法的隧道展布图拼接方法,主要解决目前隧道图像拼接中全幅拼接的技术问题。文章对一个隧道中隧道检测车分别左右检测两次所获得的两组隧道衬砌面图像进行分析,在每组数据中包含了一定量多个相机等距离触发、同一时刻同步拍摄的半幅隧道拼接数据,对隧道全幅拼接技术的进行详细分析和研究,提出了有效的解决方法,可以快速、准确地拼接出隧道展布图,满足隧道养护需求。In this paper,we study the splicing method of tunnel layout map based on FAST feature detection and FLANN feature point matching algorithm,mainly to solve the technical problem of full-frame splicing in the current tunnel image splicing.This article analyzes the two sets of tunnel lining images obtained by the tunnel inspection vehicle in a tunnel.Each set of data contains a certain amount of half-frame tunnel stitching that is triggered by multiple cameras at equal distances and shot simultaneously at the same time.Data,detailed analysis and research on the tunnel's full-frame splicing technology,and proposed effective solutions,which can quickly and accurately splice the tunnel layout to meet the needs of tunnel maintenance.

关 键 词：隧道衬砌面图像 FAST FLANN RANSAC 图像拼接 

分 类 号：U455.4[建筑科学—桥梁与隧道工程]