基于融合式PC-ORB的异源图像配准算法

伍朗; 易诗; 陈梦婷; 李立

基于融合式PC-ORB的异源图像配准算法

成都理工大学机电工程学院, 四川成都 610059

基金项目:

四川省科技厅重点研发项目 2021YFGO075

四川省科技厅重点研发项目 2021YFGO076

四川省车辆测控与安全重点实验室开放基金 OCCK2021-008

四川省重点科技项目 2020ZDZX0019

成都理工大学2021—2023年高等教育人才培养质量和教学改革项目 JG2130109

成都理工大学2021—2023年高等教育人才培养质量和教学改革项目 JG2130216

详细信息

作者简介:
伍朗（2000-），男，硕士研究生，研究方向：图像处理。E-mail: 1213836094@qq.com

通讯作者:
易诗（1983-），男，硕士生导师，研究方向：图像处理。E-mail: 549745481@qq.com

中图分类号: TN911.73
计量
- 文章访问数: 23
- HTML全文浏览量: 14
- PDF下载量: 7
- 被引次数: 0
出版历程
- 收稿日期: 2023-06-07
- 修回日期: 2023-07-11
- 刊出日期: 2024-04-20

Heterogeneous Image Registration Algorithm Based on Fusion PC-ORB

School of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu 610059, China

摘要

摘要: 异源图像配准中，由于图像的成像机理差异，图像像素强度关联和旋转畸变是不可避免的两大问题，针对图像像素强度关联问题，提出了基于辐射不变特征变换（radiation-variation insensitive feature transform，RIFT）的图像配准算法，对图像间像素关联差异小的图像对配准有良好的精度，但对旋转畸变图像会产生较多错误匹配。对于旋转畸变问题，传统的ORB（oriented fast and rotated brief）算法，对旋转图像的配准有一定的稳定性，但对于强度变化不明显的图像对，特征点检测质量较低，配准精度不理想。因此本文将相位一致性（phase consistency，PC）融合进ORB算法，利用相位信息代替传统的图像强度信息，再构造旋转不变性BRIEF特征描述子，对图像像素强度变化和旋转畸变均具有鲁棒性。用图像像素强度关联不明显的红外图像和可见光图像进行配准实验，本文算法针对不同旋转幅度的图像的配准精度较高，RMSE稳定在1.7~2.1，优于RIFT算法，在特征点检测数量、配准精度和效率等性能上均有良好性能。
- 图像配准 /
- 特征匹配 /
- 相位一致性 /
- 旋转不变性 /
- ORB算法
Abstract: In heterogeneous image registration, because of the differences in the imaging mechanisms, image pixel intensity correlation and rotation distortion are two inevitable problems. Aiming at the problem of image pixel intensity correlation, an image registration algorithm based on a radiation-invariant feature transform (RIFT) is proposed; it has good accuracy for image registration with small differences in the pixel correlation between images, but produces more error matching for rotation distortion images. For the problem of rotational distortion, the traditional Oriented Fast and Rotated Brief (ORB) algorithm has a certain degree of stability in the registration of rotating images; however, for image pairs with insignificant intensity changes, the quality of the feature point detection is low and the registration accuracy is not ideal. Therefore, this study integrates Phase Consistency into the ORB algorithm, replaces traditional image strength information with phase information, and constructs a rotation-invariant BRIEF feature descriptor that is robust to changes in the pixel strength and rotation distortion in the image. The registration experiment is conducted using infrared and visible-light images with unclear pixel intensity correlations. The algorithm proposed in this paper has high registration accuracy for images with different rotation amplitudes, and the RMSE is stable at 1.7−2.1, which is superior to the RIFT algorithm. It performs well in detecting a large number of feature points, achieving high registration accuracy, and maintaining efficiency.
- image registration /
- feature matching /
- phase consistency /
- rotational invariance /
- ORB algorithm

HTML全文

图 1 本文算法整体框架

Figure 1. Flow chart of the proposed algorithm

下载: 全尺寸图片幻灯片

图 2 可见光图像和红外图像的PC图

Figure 2. PC images of visible and infrared images

下载: 全尺寸图片幻灯片

图 3 FAST特征检测原理

Figure 3. FAST feature detection schematic diagram

下载: 全尺寸图片幻灯片

图 4 PC构造后特征点检测对比

Figure 4. Comparison of feature point detection after PC construction

下载: 全尺寸图片幻灯片

图 5 改进前后的BRIEF的对比

Figure 5. Comparison of BRIEF before and after improvement

下载: 全尺寸图片幻灯片

图 6 RANSAC算法原理示意图

Figure 6. Schematic diagram of RANSAC algorithm principle

下载: 全尺寸图片幻灯片

图 7 旋转图像的正确匹配率

Figure 7. Correct matching rate of rotated images

下载: 全尺寸图片幻灯片

图 8 旋转图像的RMSE

Figure 8. RMSE of rotated images

下载: 全尺寸图片幻灯片

图 9 旋转图像的特征点匹配结果

Figure 9. Matching results of feature points in rotating images

下载: 全尺寸图片幻灯片

图 10 不同算法的配准效果

Figure 10. Registration renderings of different algorithms

下载: 全尺寸图片幻灯片

表 1 不同算法特征匹配性能指标对比

Table 1. Comparison of performance indicators for feature matching of different algorithms

Algorithm	Number of feature points	Number of correct matches	RMSE	Registration time consumed/s
SIFT	536	/	/	1.732
ORB	248	153	2.04	0.153
RIFT	1347	1258	1.72	0.772
Ours	1425	1372	1.62	0.682

下载: 导出CSV

参考文献(16)

[1]	韩静, 柏连发, 张毅, 等. 基于改进配准测度的红外与可见光图像配准[J]. 红外技术, 2011, 33(5): 271-274. doi: 10.3969/j.issn.1001-8891.2011.05.006 HAN Jing, BAI Lianfa, ZHANG Yi, et al. Registration of infrared and visible light images based on improved registration measure[J]. Infrared Technology, 2011, 33(5): 271-274. doi: 10.3969/j.issn.1001-8891.2011.05.006
[2]	CHEN Y, ZHANG X, ZHANG Y, et al. Visible and infrared image registration based on region features and edginess[J]. Machine Vision and Applications, 2018, 29(1): 113-123. doi: 10.1007/s00138-017-0879-6
[3]	MA J, ZHAO J, MA Y, et al. Non-rigid visible and infrared face registration via regularized Gaussian fields criterion[J]. Pattern Recognition, 2015, 48(3): 772-784. doi: 10.1016/j.patcog.2014.09.005
[4]	LI Y, YU F, CAI Q, et al. Image fusion of fault detection in power system based on deep learning[J]. Cluster Computing, 2019, 22: 9435-9443. doi: 10.1007/s10586-018-2264-2
[5]	ZHUANG Y, GAO K, MIU X, et al. Infrared and visual image registration based on mutual information with a combined particle swarm optimization-powell search algorithm[J]. Optik, 2016, 127(1): 188-191. doi: 10.1016/j.ijleo.2015.09.199
[6]	Brown L G. Survey of image registration techniques[J]. ACM Computing Surveys, 1992, 24(4): 325-376. doi: 10.1145/146370.146374
[7]	Pluim J P W, Antoine Maintz J B, Viergever M A. Image registration by maximization of combined mutual information and gradient information[C]//3rd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI, 2000: 809-814.
[8]	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
[9]	WANG S, YOU H, FU K. BFSIFT: a novel method to find feature matches for sar image registration[C]//IEEE Geoscience and Remote Sensing Letters, 2012, 9(4): 649-653.
[10]	Rublee E, Rabaud V, Konolige K, et al. ORB: an efficient alternative to SIFT or SURF[C]//2011 International Conference on Computer Vision, 2011: DOI: 10.1109/ICCV.2011.6126544.
[11]	FENG Y, LI S. Research on an image mosaic algorithm based on improved ORB feature combined with SURF[C/OL]//Chinese Control and Decision Conference (CCDC), 2018: https://cpfd.cnki.com.cn/Article/CPFDTOTAL-KZJC201806004118.htm.
[12]	LI J, HU Q, AI M. RIFT: Multi-modal image matching based on radiation-variation insensitive feature transform[J]. IEEE Transactions on Image Processing, 2020, 29: 3296-3310. doi: 10.1109/TIP.2019.2959244
[13]	王珂, 邓安健, 臧文乾. 基于改进ORB和匹配策略融合的图像配准方法[J]. 测绘与空间地理信息, 2023, 46(2): 43-47. https://www.cnki.com.cn/Article/CJFDTOTAL-DBCH202302011.htm WANG Ke, DENG Anjian, ZANG Wenqian. Image registration method based on improved ORB and fusion matching strategy[J]. Geomatics and Spatial Information Technology, 2023, 46(2): 43-47. https://www.cnki.com.cn/Article/CJFDTOTAL-DBCH202302011.htm
[14]	Arrospide J, Salgado L. Log-gabor filters for image-based vehicle verification[J]. IEEE Transactions on Image Processing, 2013, 22(6): 2286-2295. doi: 10.1109/TIP.2013.2249080
[15]	WU G, ZHOU Z. An improved ORB feature extraction and matching algorithm[C]//33rd Chinese Control and Decision Conference, CCDC, 2021: DOI: 10.1109/CAC51589.2020.9327165.
[16]	孙世宇, 张岩, 胡永江, 等. 改进模型估计的无人机侦察视频快速拼接方法[J]. 红外与激光工程, 2018, 47(9): 382-390. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ201809055.htm SUN Shiyu, ZHANG Yan, HU Yongjiang, et al. Rapid stitching method for unmanned aerial vehicle reconnaissance videos based on improved model estimation[J]. Infrared and Laser Engineering, 2018, 47(9): 382-390. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ201809055.htm