共查询到20条相似文献,搜索用时 46 毫秒
1.
2.
回环检测是SLAM系统中关键部分,是减少移动机器人在运行过程中产生的累积误差的重要步骤。传统的回环检测算法使用人工建立特征,易忽略有用信息。本文提出了基于卷积神经网络(CNN)的回环检测算法,对预先训练的卷积神经网络模型(ResNet50、ResNet101、ResNet152)进行性能对比;通过改进关键帧的选取策略,将筛选出的关键帧输入预先训练好的ResNet模型,输出高维特征向量;利用主成分分析(PCA)白化来降低特征向量的维数,并通过计算特征向量间的欧式距离;最后计算相似矩阵,验证回环的准确率。通过对比实验可知,本文算法性能优于其它回环检测算法。 相似文献
3.
本文针对SLAM问题中的回环检测算法进行研究,结合深度学习技术,提出了基于Mask R-CNN的回环检测算法。该算法利用MaskR-CNN检测得到的掩膜表示图像特征,定义了一种基于物体掩膜的计算图像相似度的方法。本文还将该算法与经典的BOW算法进行了融合,提出了2种融合算法的方式。本文在2个开放的数据集NewCollege和CityCentre对提出的算法进行了实验。实验结果表明,融合算法有着比BOW算法更高的性能,而且随着图像中可识别物体数量的增加,性能也会增加。 相似文献
4.
针对存在明显光照变化或遮挡物等室外复杂场景下,现有基于深度学习的视觉即时定位与地图构建(visual simultaneous localization and mapping,视觉SLAM)回环检测方法没有很好地利用图像的语义信息、场景细节且实时性差等问题,本文提出了一种YOLO-NKLT视觉SLAM回环检测方法。采用改进损失函数的YOLOv5网络模型获取具有语义信息的图像特征,构建训练集,对网络重训练,使提取的特征更加适用于复杂场景下的回环检测。为了进一步提高闭环检测的实时性,提出了一种基于非支配排序的KLT降维方法。通过在New College数据集和光照等变化更复杂的Nordland数据集上进行实验,结果表明:室外复杂场景下,相较于其他传统和基于深度学习的方法,所提方法具有更高的鲁棒性,可以取得更佳的准确率和实时性表现。 相似文献
5.
为提高视觉同时定位与建图(SLAM)回环检测的准确率和召回率,提出一种基于改进线带描述符(LBD)和数据依赖度量的点线特征视觉SLAM回环检测算法.首先,针对现有LBD二进制转换操作只在各个条带之间比较大小而忽略条带内部属性,从而导致匹配正确率低的问题,增加了条带描述子内部对比操作.然后,考虑到视觉单词词频分布信息对相... 相似文献
6.
同步定位与建图(Simultaneous Localization and Mapping, SLAM)是移动机器人实现自主定位与导航的关键技术,已成为该领域研究的热点。视觉SLAM是指相机作为仅有的外部传感器,进行同步定位与建图的技术,随着计算机视觉的迅速发展,视觉SLAM因为信息量大、成本低廉、适用范围广和可提取语义信息等优点受到广泛关注,而回环检测(Loop Closure Detection, LCD)作为其重要的一个环节,受到学者的广泛研究。对视觉SLAM系统进行简单概述,对LCD的原理、传统的LCD算法分类和主流的LCD算法进行总结归纳,介绍了LCD的性能评估标准,对LCD当前面临的挑战及未来前景进行展望。 相似文献
7.
《信息通信》2017,(10)
即时定位与地图构建(SLAM)是解决移动机器人在未知非结构化环境中自主导航与控制的关键,一个完整的SLAM系统包括传感器数据处理、位姿估计、构建地图、回环检测四个部分。其中回环检测机制是解决移动机器人的闭环重定位,提高SLAM系统鲁棒性的重要环节。该研究提出一种基于ORB词袋模型的SLAM系统框架,通过研究与分析了使用FLANN算法选取关键帧与匹配帧间特征点,ORB特征描述子对检测速度的提高,通过k-means++算法对特征点进行训练生成含有视觉单词的词袋模型,使用高斯金字塔的直方图交叉核的SVM分类器,使用e PNP算法的增量式帧间位姿估计,回环检测重定位机制等环节,实现了单目视觉SLAM系统的初始化与位姿优化,实现了在丢帧状况下通过词袋模型进行重定位。最后通过搭建实验平台和标准数据集的测试得到的数据结果表明,基于ORB词袋模型的SLAM系统,具有良好的实时性,能够有效提高SLAM系统的重定位准确性,增强了系统的鲁棒性。 相似文献
8.
旋翼无人机能够实时精准感知自身位置是无人机实现后续相关技术的关键前提之一。为了提高旋翼无人机的定位精度,提出了一种基于激光雷达的SLAM(Simultaneous Localization and Mapping)定位方法。该方法通过融合三维激光雷达与IMU(Inertial Measurement Unit)来提升系统整体性能,对点云进行降采样,利用激光点云信息对旋翼无人机的高程进行计算,对激光雷达帧间匹配得到的有累计误差的高度变化进行更新,利用回环检测技术增加闭环约束,最后在SLAM系统后端进行联合优化。在保证无人机平稳飞行的状态下,该方法比A-LOAM算法在轨迹的平均误差上降低了约4倍,高程精度提升一个数量级至厘米级,改进了系统对高度不敏感以及误差积累过大的问题,提高了无人机工作效率及安全性。 相似文献
9.
10.
11.
In this paper,we provide a new approach for intelligent traffic transportation in the intelligent vehicular networks,which aims at collecting the ve-hicles'loca... 相似文献
12.
13.
以太网的环路检测技术 总被引:1,自引:0,他引:1
以太网在局域网中取得了巨大的成功,但是在城域网应用领域中仍需要解决网络环路的相关问题。根据不同的以太网应用领域,文章分析了几种环路检测的解决方案,包括生成树协议(STP)、以太网环路保护切换协议(ERPS)、环回检测和成环点定位技术。其中,成环点定位技术新颖实用,非常适合各种以太网局域网和城域网,对于以太网的运行和维护都有很大的意义。目前,全球的标准组织均正在积极对以太网环路检测技术进行标准化,随着标准的不断成熟,以太网的环路检测技术将逐步降低以太网的环路风险,提高以太网的可靠性,便于网络的管理。 相似文献
15.
16.
针对稀疏表示目标检测理论中稀疏度难以确定的问题,本文将联合表示应用于目标检测,提出了一种新颖的目标检测算法,并给出了该算法的非线性形式.其核心思想是:背景像元的光谱能够被其周围背景像元的光谱(背景字典)线性表示,而目标像元的光谱只能被其周围背景像元的光谱和目标先验光谱(联合字典)线性表示.该算法首先用背景字典和联合字典分别对待检测像元进行联合表示,然后比较两次联合表示的重构误差确定像元类别.通过真实的高光谱图像进行验证,结果表明,与其它目标检测算法相比,该算法具有较好的检测性能. 相似文献
17.
Chun-Hsien Wu Shiunn-Jang Chern 《Signal Processing, IEEE Transactions on》2007,55(12):5630-5642
This work devises a minimum bit error rate (BER) block-based precoder used in block transmission systems with the proposed cascaded zero-forcing (ZF) equalizer. The study framework is developed as follows. For a block-based precoder, a received signal model is formulated for the two redundancy schemes, viz., trailing-zeros (TZ) and cyclic-prefix (CP). By exploiting the property of oblique projection, a cascaded equalizer for block transmission systems is proposed and implemented with a scheme, in which the inter-block interference (IBI) is completely eliminated by the oblique projection and followed by a matrix degree-of-freedom for inter-symbol interference (ISI) equalization. With the available channel state information at the transmitter side, the matrix for ISI equalization of the cascaded equalizer is utilized to design an optimum block-based precoder, such that the BER is minimized, subject to the ISI-free and the transmission power constraints. Accordingly, the cascaded equalizer with the ISI-free constraint yields a cascaded ZF equalizer. Theoretical derivations and simulation results confirm that the proposed framework not only retains identical BER performance to previous works for cases with sufficient redundancy, but also allows their results to be extended to the cases of insufficient redundancy. 相似文献
18.
19.
《IEEE transactions on bio-medical engineering》2009,56(4):1098-1107
20.
Sensor network generally detects target at a fixed frequency. Detection interval means time spacing between two adjacent detection
attempts. While designing a sensor network for detection of target intrusion in a specific region, the interval should be
carefully set with trade-off between power consumption and detection performance. This is because redundant power may be consumed
if it is too short and the target may be missed if too long. In this paper, we study the determination of the maximum detection
interval (MDI) with specified detection performance. Path exposure is adopted as a performance metric. For detection-oriented
application, a novel method to evaluate the minimum path exposure (MPE) is developed. Then the MDI problem is formulated and
its solution is presented. The factors influencing the MDI are extensively simulated.
KeBo Deng received the B.S.E.E. degree from Nanjing University of Science and Technology, Nanjing, China, in 2003. Since 2003, he has been a Ph.D. candidate in the discipline of Communications and Information Systems at the Nanjing University of Science and Technology. His research interests mainly include sensor network and collaborative signal processing.
Zhong Liu received the B.S.E.E. degree from Anhui University, Anhui, China, in 1983, the M.S.E.E and Ph.D. degrees from University of Electronic Science and Technology of China, Chengdu, in 1986 and 1988, respectively. Since 1989, he has been a member of the faculty of the Nanjing University of Science and Technology, Nanjing, China, where he is Professor of Electronic Engineering and Dean of School of Electronic and Optoelectronic Engineering. From 1991 to 1993, he was a Postdoctoral Research Fellow at the Kyoto University, Kyoto, Japan. From 1997 to 1998, he was a visiting scholar at the Chinese University of Hong Kong, Hong Kong, China. His research interests mainly include radar signal processing, collaborative signal processing and chaotic information dynamics. 相似文献
Zhong LiuEmail: |
KeBo Deng received the B.S.E.E. degree from Nanjing University of Science and Technology, Nanjing, China, in 2003. Since 2003, he has been a Ph.D. candidate in the discipline of Communications and Information Systems at the Nanjing University of Science and Technology. His research interests mainly include sensor network and collaborative signal processing.
Zhong Liu received the B.S.E.E. degree from Anhui University, Anhui, China, in 1983, the M.S.E.E and Ph.D. degrees from University of Electronic Science and Technology of China, Chengdu, in 1986 and 1988, respectively. Since 1989, he has been a member of the faculty of the Nanjing University of Science and Technology, Nanjing, China, where he is Professor of Electronic Engineering and Dean of School of Electronic and Optoelectronic Engineering. From 1991 to 1993, he was a Postdoctoral Research Fellow at the Kyoto University, Kyoto, Japan. From 1997 to 1998, he was a visiting scholar at the Chinese University of Hong Kong, Hong Kong, China. His research interests mainly include radar signal processing, collaborative signal processing and chaotic information dynamics. 相似文献