详细信息
文献类型:期刊文献
中文题名:基于视觉里程计的单目红外视频三维重建
英文题名:3D reconstruction of monocular infrared video based on visual odometer
作者:陈博洋[1,2];孙韶媛[1,2];叶国林[1,2];赵海涛[3]
机构:[1]东华大学信息科学与技术学院,上海201620;[2]东华大学数字化纺织服装技术教育部工程研究中心,上海201620;[3]华东理工大学信息科学与工程学院,上海200237
年份:2017
卷号:47
期号:10
起止页码:1271
中文期刊名:激光与红外
外文期刊名:Laser & Infrared
收录:CSTPCD;;Scopus;北大核心:【北大核心2014】;CSCD:【CSCD2017_2018】;
基金:国家自然科学基金项目(No.61375007);上海市科委基础研究项目(No.15JC1400600)资助
语种:中文
中文关键词:红外视频;三维重建;直接法和稀疏法视觉里程计
外文关键词:infrared video;;3D reconstruction;;direct sparse odometer
摘要:针对红外视频的特点,提出了一种基于直接法和稀疏法视觉里程计的单目红外视频三维重建方法。该方法首先通过对红外热像仪标定获得热像仪内参,然后构建直接法和稀疏法视觉里程计模型,视觉里程计前端执行帧管理和点管理的任务,利用滑动窗口并借助高斯-牛顿迭代对总光度误差进行优化,计算出直接法和稀疏法视觉里程计模型所依赖的所有变量,完成定位热像仪和建图的任务。通过实验证明了该方法能够实时实现对单目红外视频进行三维重建。
Aiming at the characteristics of infrared video,a three-dimensional reconstruction method of monocular infrared video based on direct and sparse visual odometer is proposed. In this method,the internal parameters of the thermal imager are obtained by the calibration of the infrared camera,and then the direct and sparse visual odometer model is constructed. Visual odometer front-end performs frame management and point management tasks. The total photometric error is optimized by the sliding window and the Gauss-Newton iteration,and all the variables that the direct and sparse visual odometer model needs are calculated. Finally,the tasks of locating the thermal imager and plotting are fullfilled. Experimental results show that the proposed method can achieve the three-dimensional reconstruction of monocular infrared video in real time.
参考文献:
正在载入数据...
