首页|卡尔曼滤波下激光雷达扫描大数据检测算法

卡尔曼滤波下激光雷达扫描大数据检测算法

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
激光雷达扫描的不规则激光回波信号通过地面反射和空间传播，但是由于颜色较暗的粗糙物体表面反射率较小，信号易发生漫反射，大量反射信号被接收，会形成较大的接收噪声，导致激光雷达扫描大数据出现较大程度的异常。为此提出基于卡尔曼滤波的激光雷达扫描大数据检测算法。利用卡尔曼滤波算法剔除激光雷达数据中的野值;基于此，建立数据硬性约束条件，获取数据流形规则，提取激光雷达数据特征;根据提取的特征，利用支持向量回归算法完成数据分类，实现激光雷达扫描大数据检测。实验结果表明，研究方法检测激光雷达扫描大数据时，查全率始终高于 90%，查准率高于 95%，平均耗时65。1ms，且以无人小车行驶轨迹的雷达数据为测试对象时，研究方法具有较高的检测精度，说明该方法的研究价值较高，应用效果更好。
Kalman Filtering Based Big Data Detection of Lidar Scanning
Irregular laser echo signal scanned by lidar is reflected on the ground and propagated in space.Due to the low reflectivity of the surface of dark rough object,the signal is prone to diffuse reflection.And the reflected sig-nals are received,thus forming a large receiving noise and leading to a large degree of abnormality in the big data scanned by lidar.Therefore,an algorithm for detecting lidar scanning big data was proposed based on Kalman filter.The Kalman filtering algorithm was used to eliminate the outliers in lidar data at first.On this basis,the rigid con-straints of data were established,and then the data manifold rules were obtained to extract the lidar data features.Ac-cording to the extracted features,support vector regression algorithm was adopted to complete the data classification,thus realizing the detection of lidar scanning big data.Following conclusions can be drawn from experimental results.The recall rate is always more than 90%,the precision rate is higher than 95%,and the average time is 65.1ms when the proposed method is used to detect the big data scanned by the lidar.When the radar data of the driving track of the unmanned car is used as the test object,the method has a high detection accuracy,indicating that the research val-ue of the method is higher and the application effect is better.

Kalman filteringLidarBig data detectionData outliersFeature extractionSupport vector regres-sion algorithm

郑冬花、邓铭毅、叶丽珠

展开 >

广州商学院信息技术与工程学院,广东 广州 511363

广东工业大学,广东 广州 511300

卡尔曼滤波 激光雷达 大数据检测 数据野值 特征提取 支持向量回归算法

广东省2020年本科高校教学质量与教学改革工程建设特色专业项目广东省高等教育学会"十四五"规划2021年度高等教育研究课题广东省教育厅普通高校认定类科研项目

粤教高函[2020]19号-7621GYB082021KTSCX150

2024

计算机仿真
中国航天科工集团公司第十七研究所

计算机仿真

CSTPCD
影响因子：0.518
ISSN：1006-9348
年,卷(期)：2024.41(6)
  • 11