Static image registration algorithm in high-precision video registration algorithm

Ping WANG

doi:10.3788/YJYXS20203506.0612

您当前的位置：

首页 >

文章列表页 >

Static image registration algorithm in high-precision video registration algorithm

Image Processing | 更新时间：2020-07-13

- Static image registration algorithm in high-precision video registration algorithm
- Chinese Journal of Liquid Crystals and Displays Vol. 35, Issue 6, Pages: 612-618(2020)
- 作者机构：
  
  1.阳光学院人工智能学院, 福建福州 350015
  2.空间数据挖掘与应用福建省高校工程研究中心, 福建福州 350015
- 作者简介：
- 基金信息：
  
  Basic Scientific Research Project of Fujian Provincial Public Welfare Scientific Research Institute in 2019(2019R1008-9)
- DOI：10.3788/YJYXS20203506.0612
  CLC： TP391.41
- Received：21 December 2019，
  
  Revised：14 February 2020，
  
  Accepted：14 February 2020，
  
  Published：05 June 2020
- 稿件说明：
移动端阅览
Ping WANG. Static image registration algorithm in high-precision video registration algorithm[J]. Chinese journal of liquid crystals and displays, 2020, 35(6): 612-618.
DOI：

Ping WANG. Static image registration algorithm in high-precision video registration algorithm[J]. Chinese journal of liquid crystals and displays, 2020, 35(6): 612-618. DOI： 10.3788/YJYXS20203506.0612.

摘要

为优化视频配准工作中的静态图像配准算法，本文应用深度学习的卷积神经网络VGGNet设计了一个静态图像配准算法。通过仿真实验结果可知，VGGNet能够被很好地应用于静态图像配准工作中；结合均方根误差和

red

结果可知，基于Conv5、FC1或者FC2输出特征的静态图像配准算法具有良好的性能，基于FC2的配准算法性能最好，其均方根误差值为0.031 22，

red

值为74。通过视频缩放和亮度转换后，静态图像配准算法的性能有所下降，在亮度变换后基于FC2输出特征的配准算法性能较高。与传统HOG、LBP特征提取的图像配准算法相比，本文算法具备较好的配准精度和正确率。研究结果可为当前基于深度学习卷积神经网络的静态图像配准算法提供参考。

Abstract

In order to optimize the static image registration algorithm in video registration

a static image registration algorithm based on deep learning convolutional neural network VGGNet is proposed. The simulation experiments results show that VGGNet can be well applied to the static image registration work. Combining the root mean square error and

red

results

it can be seen that the static image registration algorithm based on the output features of Conv5

FC1

or FC2 has good performance. The algorithm based on FC2 has the best performance with root mean square error value of 0.031 22 and

red

of 74. After video scaling and brightness conversion

the performance of the static image registration algorithm decreases. The performance of the registration algorithm based on FC2 is better overall after the brightness conversion. Compared with the traditional image registration algorithm based on HOG and LBP feature extraction

this algorithm has better registration precision and accuracy. The research results can provide a reference for current static image registration algorithms based on deep learning convolutional neural networks.

关键词

Keywords

references

熊平 , 吕顺华 . 多视频多图像实时配准 . 计算机应用与软件 , 2014 . 31 ( 5 ): 180 - 183 . http://d.old.wanfangdata.com.cn/Periodical/jsjyyyrj201405047 http://d.old.wanfangdata.com.cn/Periodical/jsjyyyrj201405047 .

P XIONG , S H LYU . Multi video and multi image real-time registration . Computer application and software , 2014 . 31 ( 5 ): 180 - 183 . http://d.old.wanfangdata.com.cn/Periodical/jsjyyyrj201405047 http://d.old.wanfangdata.com.cn/Periodical/jsjyyyrj201405047 .

任思阳.基于DSP的多源视频配准算法研究[D].长春: 长春理工大学, 2018.

REN S Y. Research on multi-source video registration algorithm based on DSP[D]. Changchun: Changchun University of technology, 2018..(in Chinese)

杨茹.基于时空相位一致性的多传感器视频配准算法研究与实现[D].西安: 西安电子科技大学, 2013.

YANG R. Research and implementation of multi-sensor video registration algorithm based on temporal and spatial phase consistency[D]. Xi'an: Xi'an University of Electronic Science and technology, 2013..(in Chinese)

Y CASPI , D SIMAKOV , M IRANI . Feature-based sequence-to-sequence matching . International Journal of Computer Vision , 2006 . 68 ( 1 ): 109 - 116 . http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=ddd90639a00c91b243d597496c7cfd0e http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=ddd90639a00c91b243d597496c7cfd0e .

SHAKIL O. An efficient video alignment approach for non-overlapping sequences with free camera movement[C]// Acoustics, Speech and Signal Processing 2006, Toulouse, France, IEEE, 2006: Ⅱ257-Ⅱ260.

P SAND , S TELLER . Video matching . ACM Transactions on Graphics , 2004 . 22 ( 3 ): 592 - 598 . http://d.old.wanfangdata.com.cn/Periodical/cqydxyxb-zrkx201503020 http://d.old.wanfangdata.com.cn/Periodical/cqydxyxb-zrkx201503020 .

陈为龙 , 何小海 , 郭黎 , 等 . 高精度多视频配准算法中静态图像配准算法的选取 . 四川大学学报:自然科学版 , 2012 . 49 ( 5 ): 1053 - 1059 . http://d.old.wanfangdata.com.cn/Periodical/scdxxb201205021 http://d.old.wanfangdata.com.cn/Periodical/scdxxb201205021 .

W L CHEN , X H HE , L GUO , 等 . Selection of static image registration algorithm in high-precision multi video registration algorithm . Journal of Sichuan University:Natural Science Edition , 2012 . 49 ( 5 ): 1053 - 1059 . http://d.old.wanfangdata.com.cn/Periodical/scdxxb201205021 http://d.old.wanfangdata.com.cn/Periodical/scdxxb201205021 .

茆旋宇 , 郑子扬 , 王佩 , 等 . 基于深度卷积网络的PRI调制模式识别方法 . 航天电子对抗 , 2019 . 35 ( 5 ): 32 - 37 . http://d.old.wanfangdata.com.cn/Periodical/htdzdk201905009 http://d.old.wanfangdata.com.cn/Periodical/htdzdk201905009 .

X Y MAO , Z Y ZHENG , P WANG , 等 . PRI modulation pattern recognition method based on deep convolution network . Aerospace electronic warfare , 2019 . 35 ( 5 ): 32 - 37 . http://d.old.wanfangdata.com.cn/Periodical/htdzdk201905009 http://d.old.wanfangdata.com.cn/Periodical/htdzdk201905009 .

根绒切机多吉.基于粗糙集的CNN图像识别算法研究[D].成都: 成都理工大学, 2019.

GENRONGQIEJIDUOJI. Research on CNN image recognition algorithm based on rough set[D]. Chengdu: Chengdu University of technology, 2019.(in Chinese)

吕朦.基于多尺度卷积神经网络的图像分类算法研究[D].武汉: 湖北工业大学, 2019.

LYU M. Image classification algorithm based on multi-scale convolution neural network[D]. Wuhan: Hubei University of technology, 2019.

赵子豪.基于卷积神经网络的白细胞图像检测及分类技术[D].西安: 西北大学, 2019.

ZHAO Z H. Detection and classification of leukocyte image based on convolution neural network[D]. Xi'an: Northwest University, 2019..(in Chinese)

马得花 , 闫宏亮 . 全卷积神经网络的车辆点云目标精细化检测 . 测绘科学 , 2020 . ( 2 ): 1 - 9 . http://d.old.wanfangdata.com.cn/Periodical/chkx202003015 http://d.old.wanfangdata.com.cn/Periodical/chkx202003015 .

D H MA , H L YAN . Vehicle point cloud target fine detection based on full convolution neural network . Mapping Science , 2020 . ( 2 ): 1 - 9 . http://d.old.wanfangdata.com.cn/Periodical/chkx202003015 http://d.old.wanfangdata.com.cn/Periodical/chkx202003015 .

刘淑萍 , 刘羽 , 於俊 , 等 . 结合手指检测和HOG特征的分层静态手势识别 . 中国图象图形学报 , 2015 . 20 ( 6 ): 781 - 788 . http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201506007 http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201506007 .

S P LIU , Y LIU , J YU , 等 . Hierarchical static gesture recognition combined with finger detection and hog features . Chinese Journal of image graphics , 2015 . 20 ( 6 ): 781 - 788 . http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201506007 http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201506007 .

周汐 , 曹林 . 分块LBP的素描人脸识别 . 中国图象图形学报 , 2015 . 20 ( 1 ): 50 - 58 . http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201501006 http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201501006 .

X ZHOU , L CAO . sketch face recognition based on block LBP . Chinese Journal of image graphics , 2015 . 20 ( 1 ): 50 - 58 . http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201501006 http://d.old.wanfangdata.com.cn/Periodical/zgtxtxxb-a201501006 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Research progress on absolute visual localization of unmanned aerial vehicles

Robust stitching method for borehole wall images under multi-interference imaging conditions

3D hand reconstruction method based on adaptive occlusion recovery and topology-pose bidirectional perception

Reconstruction method of video snapshot compressive imaging based on Mamba-2

3D light-field display with corrected perspective based on optical flow prediction

Related Author

MIAO Zongcheng

ZHOU Runxi

LIU Jie

YANG Haiyan

LUO Qi

XU Wangming

LI Yaoxiang

ZHAO Yuhang

Related Institution

Technological Institute of Materials & Energy Science （TIMES）， School of Electronic Information， Xijing University

School of Artificial Intelligence， Optics and Electronics （iOPEN）， Northwestern Polytechnical University

School of Information Science and Engineering， Wuhan University of Science and Technology

School of Automation， Nanjing University of Information Science & Technology

School of Computer and Computing Science， Hangzhou City University

AI问答

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China 130033 Postal code：130033
Tel：0431-86176059 Email：yjxs@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰