基于顶点向量角的改进型扫描线滤波算法 
下载： 828次
Improved Scan Line Filtering Algorithm Based on Vertex Vector Angle
长沙理工大学交通运输学院测绘工程系, 湖南 长沙 410000
图 & 表
图 1. 顶点向量角示意图
Fig. 1. Schematic of vertex vector angle
下载图片 查看原文
图 2. 扫描线的划分结果
Fig. 2. Division result of scan line
下载图片 查看原文
图 3. 顶点向量角-扫描线滤波算法的流程示意图
Fig. 3. Flow chart of vertex vector angle-scan line filtering algorithm
下载图片 查看原文
图 4. samp12数据集中xmin+20.0<x<xmin+20.5的点云数据在不同算法处理后的滤波结果。(a)滤波前;(b)传统算法;(c)传统算法减小阈值;(d)所提算法
Fig. 4. Filtering results of point cloud data in samp12 dataset xmin+20.0<x<xmin+20.5 after processing with different algorithms. (a) Before filtering; (b) traditional algorithm; (c) traditional algorithm reduces threshold; (d) proposed algorithm
下载图片 查看原文
图 5. samp12数据集xmin+15.0<x<xmin+15.5的点云数据在不同算法处理后的滤波结果。(a)滤波前;(b)传统算法;(c)所提算法
Fig. 5. Filtering results of point cloud data in samp12 dataset xmin+15.0<x<xmin+15.5 after processing with different algorithms. (a) Before filtering; (b) traditional algorithm; (c) proposed algorithm
下载图片 查看原文
图 6. 传统算法滤波后的点云分布
Fig. 6. Point cloud distribution after traditional algorithm filtering
下载图片 查看原文
图 7. 所提算法滤波后的点云分布
Fig. 7. Point cloud distribution after proposed algorithm filtering
下载图片 查看原文
表 1samp12数据集中xmin+20.0<x<xmin+20.5在不同算法处理后的误差
Table1. Errors of xmin+20.0<x<xmin+20.5 in samp12 dataset after processing by different algorithms
| Algorithm | Number of sample points | Number ofground points | Number offeatures | Type Ierror /% | Type IIerror/% | Totalerror/% |
|---|
| Traditional algorithm | 137 | 61 | 76 | 13.43 | 4.29 | 8.76 | | Traditional algorithm reduces threshold | 137 | 47 | 90 | 31.34 | 1.43 | 16.06 | | Proposed algorithm | 137 | 59 | 78 | 13.43 | 1.43 | 7.30 |
|
查看原文
表 2samp12数据集中xmin+15.0<x<xmin+15.5在不同算法处理后的误差统计
Table2. Errors of xmin+15.0<x<xmin+15.5 in samp12 dataset after processing by different algorithms
| Algorithm | Number of sample points | Number ofground points | Number offeatures | Type Ierror/% | Type IIerror/% | Totalerror/% |
|---|
| Traditional algorithm | 134 | 48 | 86 | 8.51 | 5.75 | 6.72 | | Proposed algorithm | 134 | 43 | 91 | 8.51 | 0 | 2.99 |
|
查看原文
表 3不同算法处理不同数据集的误差统计
Table3. Error statistics of different algorithms processing different datasets
| Sample | Algorithm | Number of sample points | Number ofground points | Number offeatures | Type Ierror/% | Type IIerror/% | Totalerror/% |
|---|
| Samp12 | Sample point | 52119 | 26691 | 25428 | | | | | Traditional algorithm | 52119 | 24339 | 27580 | 17.52 | 9.93 | 13.82 | | Proposed algorithm | 52119 | 23741 | 27378 | 17.61 | 6.89 | 12.38 | | Samp21 | Sample point | 12960 | 10085 | 2875 | | | | | Traditional algorithm | 12960 | 9983 | 2977 | 7.66 | 23.30 | 11.14 | | Proposed algorithm | 12960 | 9703 | 3257 | 7.74 | 13.90 | 9.10 | | Samp31 | Sample point | 28862 | 15556 | 13306 | | | | | Traditional algorithm | 28862 | 15928 | 12934 | 6.51 | 10.41 | 8.31 | | Proposed algorithm | 28862 | 15177 | 13685 | 6.79 | 5.09 | 6.00 |
|
查看原文
蒋志祥, 邓兴升, 邢承滨. 基于顶点向量角的改进型扫描线滤波算法[J]. 激光与光电子学进展, 2020, 57(23): 232801. Zhixiang Jiang, Xingsheng Deng, Chengbin Xing. Improved Scan Line Filtering Algorithm Based on Vertex Vector Angle[J]. Laser & Optoelectronics Progress, 2020, 57(23): 232801.
PDF全文
