一种基于区域划分的改进ORB算法

孙浩; 王朋

doi:10.13700/j.bh.1001-5965.2020.0054

一种基于区域划分的改进ORB算法

doi: 10.13700/j.bh.1001-5965.2020.0054

孙浩,
王朋^,

山东交通学院信息科学与电气工程学院, 济南 250357

基金项目:

国家自然科学基金 61502277

山东省自然科学基金 ZR2015FL018

详细信息

作者简介:
孙浩男, 硕士研究生。主要研究方向:图像特征提取与匹配

王朋男, 博士, 副教授, 硕士生导师。主要研究方向:数字图像处理、视频编码与传输控制、计算机视觉

通讯作者:
王朋, E-mail:knightwp@126.com

中图分类号: TP391.4
计量
- 文章访问数: 1045
- HTML全文浏览量: 225
- PDF下载量: 111
- 被引次数: 0
出版历程
- 收稿日期: 2020-02-27
- 录用日期: 2020-04-18
- 网络出版日期: 2020-09-20

An improved ORB algorithm based on region division

SUN Hao,
WANG Peng^,

School of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, China

Funds:

National Natural Science Foundation of China 61502277

Natural Science Foundation of Shandong Province ZR2015FL018

More Information

Corresponding author: WANG Peng, E-mail:knightwp@126.com

摘要

摘要:
针对传统ORB算法所提取的特征点分布不均匀、存在冗杂，且不具有尺度不变性的问题，提出了一种基于区域划分的改进ORB算法。算法根据需要提取的特征点总数和所划分的区域个数计算每个小区域需要提取的特征点个数，解决了在特征点提取过程中特征点重叠和特征点冗余的问题；通过构建图像金字塔，在每一层图像金字塔上提取特征点，解决了ORB算法提取的特征点不具有尺度不变性的问题。实验结果表明：在不损失图像匹配精度的同时，所提算法提取的特征点更加均匀合理，在提取速度上也较传统ORB算法提升了16%左右。
- ORB算法 /
- 特征点提取 /
- 区域划分 /
- 图像金字塔 /
- 图像匹配
Abstract:
The feature points extracted by the traditional ORB algorithm are not evenly distributed, are redundant and have no scale invariance. To solve this problem, this paper proposes an improved ORB algorithm based on region division. According to the total number of feature points to be extracted and the number of regions to be divided, the algorithm calculates the number of feature points to be extracted for each region, which solves the problem of feature point overlap and redundancy in the feature point extraction process. By constructing the image pyramid and extracting feature points on each layer, the problem that the feature points extracted by ORB algorithm do not have scale invariance is solved. The experimental results show that the feature points extracted by our algorithm are more uniform and reasonable without losing the accuracy of image matching, and the extraction speed is about 16% faster than that of the traditional ORB algorithm.
- ORB algorithm /
- feature point extraction /
- region division /
- image pyramid /
- image matching

HTML全文

图 1 FAST特征点提取示意图

Figure 1. Schematic diagram of FAST feature point extraction

下载: 全尺寸图片幻灯片

图 2 描述子计算示意图

Figure 2. Schematic diagram of descriptor calculation

下载: 全尺寸图片幻灯片

图 3 改进ORB算法的流程图

Figure 3. Flow chart of improved ORB algorithm

下载: 全尺寸图片幻灯片

图 4 图像金字塔示意图

Figure 4. Image pyramid illustration

下载: 全尺寸图片幻灯片

图 5 原始图像

Figure 5. Original images

下载: 全尺寸图片幻灯片

图 6 不同算法提取结果

Figure 6. Extraction results of different algorithms

下载: 全尺寸图片幻灯片

图 7 不同算法匹配结果

Figure 7. Matching results of different algorithms

下载: 全尺寸图片幻灯片

图 8 不同算法运行截图

Figure 8. Screenshot of running of different algorithms

下载: 全尺寸图片幻灯片

表 1 匹配精度对比

Table 1. Matching accuracy comparison

算法	总提取点数	正确匹配点数	匹配准确率/%
ORB算法	500	339	67.8
本文算法	469	329	70.1
文献[12]改进的ORB算法	253	233	92.1

下载: 导出CSV

表 2 算法运行时间比较

Table 2. Comparison of algorithm running time

算法	总提取点数	总运行时间/ms	平均运行时间/ms
ORB算法	500	32.7	0.065 4
本文算法	469	25.8	0.055 0
文献[11]改进的ORB算法	262	32.36	0.123 5
文献[12]改进的ORB算法	253	88.37	0.349 3

下载: 导出CSV

表 3 算法匹配时间对比

Table 3. Comparison of matching time

算法	总匹配点数	总匹配时间/ms	平均匹配时间/ms
ORB算法	339	1.50	0.004 4
本文算法	329	1.31	0.004 0

下载: 导出CSV

参考文献(18)

[1]	高翔, 张涛.视觉SLAM十四讲:从理论到实践[M].北京:电子工业出版社, 2017:10-14. GAO X, ZHANG T.14 lectures on visual SLAM:From theory to practice[M].Beijing:Publishing House of Electronics Industry, 2017:10-14(in Chinese).
[2]	张全, 盛妍, 吴佐平, 等.公共区域监控视频数据目标特征跟踪定位方法[J].自动化与仪器仪表, 2020(1):51-54. ZHANG Q, SHENG Y, WU Z P, et al.Target tracking and location method for common area surveillance video data[J].Automation & Instrumentation, 2020(1):51-54(in Chinese).
[3]	LOWE D G.Distinctive image features from scale invariant keypoints[J].International Journal of Computer Vision, 2004, 60(2):91-110. doi: 10.1023/B:VISI.0000029664.99615.94
[4]	BAY H, TUYTELAARS T, VAN GOOL L.Surf: Speeded up robust features[C]//European Conference on Computer Vision.Berlin: Springer, 2006: 404-417.
[5]	ROSTEN E, PORTER R, DRUMMOND T.Faster and better:A machine learing approach to corer detection[J].Analysis and Machine Intelligence, 2008, 32(1):105-119.
[6]	RUBLEE E, RABAUD V, KONOLIGE K, et al.ORB: An efficient alternative to SIFT or SURF[C]//Proceedings of IEEE International Conference on Computer Vision.Piscataway: IEEE Press, 2011: 2564-2571.
[7]	VISWANATHAN D G.Features from accelerated segment test(FAST)[EB/OL].(2016-04-15)[2020-01-10].
[8]	CALONDER M, LEPETIT V, STRECHA C, et al.Brief: Binary robust independent elementary features[C]//European Conference on Computer Vision.Berlin: Springer, 2010: 778-792.
[9]	刘伟, 钱莉.基于OpenCV环境的SIFT、SURF、ORB算法比较分析[J].化工自动化及仪表, 2018, 45(9):714-716. LIU W, QIAN L.Comparative analysis of SIFT and SURF and ORB algorithms based on OpenCV environment[J].Control and Instruments in Chemical Industry, 2018, 45(9):714-716(in Chinese).
[10]	ZHAO H Y, ZHAO H C, LV J F, et al.Multimodal image matching based on multimodality robust line segment descriptor[J].Neurocomputing, 2016, 177:290-303. doi: 10.1016/j.neucom.2015.11.025
[11]	潘盼.一种改进的ORB算法[J].微型机与应用, 2017, 36(20):23-26. PAN P.An improved ORB algorithm[J].Microcomputer & Its Applications, 2017, 36(20):23-26(in Chinese).
[12]	马丹, 赖惠成.改进ORB算法的图像匹配[J].计算机仿真, 2018, 35(10):274-278. MA D, LAI H C.Image matching based on improved ORB[J].Computer Simulation, 2018, 35(10):274-278(in Chinese).
[13]	陈思聪, 刘晶红, 何林阳, 等.一种用于图像拼接的改进BRISK算法[J].液晶与显示, 2016, 31(3):324-330. CHEN S C, LIU J H, HE L Y, et al.Improved BRISK algorithm used in image mosaic[J].Chinese Journal of Liquid Crystals and Displays, 2016, 31(3):324-330(in Chinese).
[14]	范新南, 顾亚飞, 倪建军.改进ORB算法在图像匹配中的应用[J].计算机与现代化, 2019(2):1-6. FAN X N, GU Y F, NI J J.Application of improved ORB algorithm in image matching[J].Computer and Modernization, 2019(2):1-6(in Chinese).
[15]	HUANG C, ZHOU W D.Fast scene matching navigation algorithm based on ORB[J].Journal of Information and Computational Science, 2014, 11(11):3857-3863.
[16]	李玉峰, 李广泽, 谷绍湖, 等.基于区域分块与尺度不变特征变换的图像拼接算法[J].光学精密工程, 2016, 24(5):1197-1205. LI Y F, LI G Z, GU S H, et al.Image mosaic algorithm based on area blocking and SIFT[J].Optics and Precision Engineering, 2016, 24(5):1197-1205(in Chinese).
[17]	王健, 于鸣, 任洪娥.一种用于图像拼接的改进ORB算法[J].液晶与显示, 2018, 33(6):520-527. WANG J, YU M, REN H E.Improved ORB algorithm used in image stitching[J].Chinese Journal of Liquid Crystals and Displays, 2018, 33(6):520-527(in Chinese).
[18]	姚海芳, 郭宝龙.一种基于ORB的特征匹配算法[J].电子设计工程, 2019, 27(16):175-179. YAO H F, GUO B L.An ORB-based feature matching algorithm[J].Electronic Design Engineering, 2019, 27(16):175-179(in Chinese).