BFSIFT: A Novel Method to Find Feature Matches for SAR Image Registration

Shan-hu, Wang; You, Hongjian; Fu, Kun

doi:10.1109/lgrs.2011.2177437

Cited by 167 publications

(33 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fan et al [11] improved the matching performance by skipping the dominant orientation assignment when the matching images do not have rotation transformation. By replacing the Gaussian filter with several anisotropic filters, the Bilateral Filter SIFT (BFSIFT), Adapted Anisotropic Gaussian SIFT (AAG-SIFT) and Nonlinear Diffusion Scale Space-SIFT (NDSS-SIFT) are proposed to improve the matching performance [12,13,14]. Dellinger et al [9] proposed the SAR-SIFT algorithm specifically dedicated to SAR images by utilizing the Ratio Of the Exponentially-Weighted Averages (ROEWA) instead of a differential to calculate a gradient.…”

Section: Introductionmentioning

confidence: 99%

An Automatic and Novel SAR Image Registration Algorithm: A Case Study of the Chinese GF-3 Satellite

Xiang

Wang

You

2018

Sensors

View full text Add to dashboard Cite

The Chinese GF-3 satellite launched in August 2016 is a Synthetic Aperture Radar (SAR) satellite that has the largest number of imaging modes in the world. It achieves a free switch in the spotlight, stripmap, scanSAR, wave, global observation and other imaging modes. In order to further utilize GF-3 SAR images, an automatic and fast image registration procedure needs to be done. In this paper, we propose a novel image registration technique for GF-3 images of different imaging modes. The proposed algorithm consists of two stages: coarse registration and fine registration. In the first stage, we combine an adaptive sampling method with the SAR-SIFT algorithm to efficiently eliminate obvious translation, rotation and scale differences between the reference and sensed images. In the second stage, uniformly-distributed control points are extracted, then the fast normalized cross-correlation of an improved phase congruency model is utilized as a new similarity metric to match the reference image and the coarse-registered image in a local search region. Moreover, a selection strategy is used to remove outliers. Experimental results on several GF-3 SAR images of different imaging modes show that the proposed algorithm gives a robust, efficient and precise registration performance, compared with other state-of-the-art algorithms for SAR image registration.

show abstract

Section: Introductionmentioning

confidence: 99%

An Automatic and Novel SAR Image Registration Algorithm: A Case Study of the Chinese GF-3 Satellite

Xiang

Wang

You

2018

Sensors

View full text Add to dashboard Cite

show abstract

“…For instance, many feature-based approaches, such moment feature, features derived from Scale-Invariant Feature Transform (SIFT) and Markov Random Field model, etc., have been proposed to register the affine geometric distortion between two optical satellite images, such as multispectral, hyperspectral and aerial images (Dai and Khorram 1999;Flusser and Suk 1994;Kasetkasem et al 2013;Li et al 2009;Wong and Clausi 2007). While most of these methods were developed for two optical images co-registration, in recent years, many approaches were also proposed to co-register two SAR images (Chen, Chen, and Su 2014;Schmitt et al 2013;Schwind et al 2010;Wang, You, and Fu 2012). Among these approaches, the SIFT method has been improved and applied towards the registration between two SAR images (Schwind et al 2010;Wang, You, and Fu 2012).…”

Section: Introductionmentioning

confidence: 99%

An automatic co-registration approach for optical and SAR data in urban areas

Zhang

Lin

2016

Annals of GIS

View full text Add to dashboard Cite

With the increasing applications of multisource remote sensing, the synergistic use of optical and Synthetic Aperture Radar (SAR) remote sensing images is becoming increasingly popular. Although numerous approaches have been proposed to register satellite images between two optical images or between two SAR images, these approaches cannot be directly applied to register optical and SAR images due to the differences of imaging mechanism between optical and SAR remote sensing. Therefore, the co-registration between optical and SAR images remains challenging, especially in urban areas where surface geometric configuration is much more complex. This study aims to propose a fast and automatic approach for optical and SAR data co-registration based on an assumption of affine transformation between the two images. With a set of initial homologous points generated automatically, an optimization is applied towards the initial homologous points on the SAR images subject to an affine transformation model and a given threshold of root mean square errors (RMSEs) for the co-registration. Three various data sets of optical and SAR images, including SPOT-5 and ENVISAT ASAR, Landsat TM and ENVISAT ASAR, and SPOT-5 and ALOS PALSAR, were employed to test and validate the effectiveness of the proposed method. Experimental results demonstrate the promising effectiveness of the proposed approach with a significant reduction of RMSE compared with the initial homologous points. The proposed approach is especially applicable in urban areas where the uncertain errors of homologous points in SAR images are higher due to the speckles and foreshortening, layover and shadows from tall buildings.

show abstract

“…Mean time, Suri combined the SIFT algorithm with mutual information [8,9]; Shanhu Wang and Hongjian You proposed a algorithm that match images roughly to eliminate the effects of image rotation and scale, then realize precise registrationusing Harris SAR image feature extraction algorithm [10] in 2012; Baoshang Zhang, Tian Zheng and Weidong Yan proposed a SAR image registration method based on the divided regions [11]; Guiqin Xia combined the corner match with prior information [12] in 2014. Registration method based on the point feature is simple and convenient.…”

Section: Introductionmentioning

confidence: 99%

A Method Based on Prior Information of Feature Points Registration for SAR Images

Li¹,

Ping-yan²,

Tan³

2016

Proceedings of the 2016 5th International Conference on Advanced Materials and Computer Science

View full text Add to dashboard Cite

Abstract. The concept of moving target detection is monitoring and detecting the moving target in the target area, achieving the goal of monitoring the target area all-day and all-weather. During the process of monitoring, the multi-frame SAR images which get from the same scenario need to be registered. The Range Doppler equations are used to obtain the location model of SAR images in this paper, combing the SIFT algorithm with the prior information. Firstly, the SIFT is used to extract the feature points. Secondly, the prior information is used to achieve the regional division of the feature points. Finally, the nearest neighbor matching rule is used to match the targets in the candidate target region. The prior information of SAR images is fully used in this method. In this way, the matching search range is reduced and the efficiency of matching algorithm is improved. Real data simulations prove that this algorithm is feasible. 1.IntroductionSAR imaging technique is used for imaging stationary targets, and GMTI technique is used for detecting and tracking moving targets. With these two technologies, SAR-GMTI system cannot only image for regional targets, but also detect moving target on the ground and form the target track. The realization of SAR-GMIT requires Sequence registration of multiple frames of SAR images in the same scene [1]. Currently, SAR image registration [2] is divided into two methods: gray information-based and features-based. Registration based on features extracts the image feature information to describe the characteristics. Then finds match feature using matching criteria and search methods to complete image registration [3].Lowe proposed SIFT feature point matching algorithm [4,5] Registration method based on the point feature is simple and convenient. However, speckle noise in SAR image can lead to a lot of feature points extracted and it will increase the amount of arithmetic operations. How to reduce the computation and improve the robustness of the algorithm is a key research direction of registration algorithm based on the feature point. This article combines SIFT algorithm with the prior information. It divides the feature points into different region using the prior information to avoid the global feature matching. As a result, the SAR image registration algorithm efficiency is improved. 2.SIFT algorithmThrough a Gaussian differential function, the SIFT algorithm generates a scale space, and selects local extreme point as the candidate feature point, then removes the unstable low contrast and edge response points, and precisely positions feature points [13] [14]. The flowchart of SIFT algorithm for image registrations shown in Fig 1.

show abstract

BFSIFT: A Novel Method to Find Feature Matches for SAR Image Registration

Cited by 167 publications

References 18 publications

An Automatic and Novel SAR Image Registration Algorithm: A Case Study of the Chinese GF-3 Satellite

An Automatic and Novel SAR Image Registration Algorithm: A Case Study of the Chinese GF-3 Satellite

An automatic co-registration approach for optical and SAR data in urban areas

A Method Based on Prior Information of Feature Points Registration for SAR Images

Contact Info

Product

Resources

About