嵌入式平台上的三维重建算法研究  

Research on 3D Reconstruction Algorithm Based on Embedded Platform

作　　者：蒋林[1] 郭晨[1] 朱志超[1] 程文凯[1] JIANG Lin, GUO Chen, ZHU Zhi-ehao, CHENG Wen-kai(College of Machinery and Automation, Wuhan Univers ity of Science and Technology, Hubei Wuhan 430081, Chin)

机构地区：[1]武汉科技大学机械自动化学院,湖北武汉430081

出　　处：《机械设计与制造》2018年第8期256-258,共3页Machinery Design & Manufacture

基　　金：国家自然科学基金资助项目(61105086);国家重点实验室开放基金资助项目(SKLRS-2010-12)

摘　　要：针对当前在嵌入式平台上的三维重建实时性和点云配准精度下降的问题,提出了一种简单的位姿估计方法并在此基础上实现了基于视觉的三维重建。将差动式驱动的移动机器人运动简化为在同一时刻只有水平转动或者同一时刻只有直线平移。对分解后的两种运动形式采用不同的简化算法,在降低计算量的同时保证模型精度。通过基于标准数据集的离线实验和基于真实环境的机器人在线实现结果表明,提出的方法能够满足室内环境重建的精度和实时性。bTo solvethe problem of real-time performance and low accuracy of point cloud registration for 3 D reconstruction on the embedded platform,this paper presents a simple pose estimation method. On this basis,completes three-dimensional reconstruction. The movement of the differential drive mobile robot is simplified to horizontal rotation or linear translation at a time. After the decomposition of the two different forms of motion using a simplified algorithm,in reducing the amount of calculation to ensure the accuracy of the model. Based on the off-line experiment of standard data and the online implementation of robot in the real environment,the result shows that the proposed method can meet the precision and realtime performance of indoor environment reconstruction.

关 键 词：嵌入式平台 运动分解 基础矩阵 奇异值分解 转换矩阵 LAPLACIAN算子 

分 类 号：TH16[机械工程—机械制造及自动化]