Hierarchical approach for synthetic aperture radar and optical image coregistration using local and global geometric relationship of invariant features

Geometrical Matching of SAR and Optical Images Utilizing ASIFT Features for SAR-based Navigation Aided Systems

“…The process of image registration refers to the alignment of two or more images of the same scene which might be obtained with the same sensor, time, and imaging conditions, as well as by different sensors and viewpoints. The process of orientating optical and SAR images is still an open issue, which creates many challenges [9][10][11][12][13][14][15][16][17][18][19]. Many issues can be dealt with in the process of the synergy of optical and SAR data, due to the great differences between passive and active remote sensing techniques.…”

“…Many issues can be dealt with in the process of the synergy of optical and SAR data, due to the great differences between passive and active remote sensing techniques. One of the issues is the problem with the speckle noise that influences the detection of robust corresponding/tie points [9,[11][12][13]15,16,18,19]. Another issue is related to the differences in the image geometry acquired from these two devices [9,[11][12][13]15,16,18,19].…”

“…One of the issues is the problem with the speckle noise that influences the detection of robust corresponding/tie points [9,[11][12][13]15,16,18,19]. Another issue is related to the differences in the image geometry acquired from these two devices [9,[11][12][13]15,16,18,19]. Thus, many co-registration approaches have been proposed (e.g., [9][10][11][12][13][14][15][16][17][18][19]), but, in general, these approaches might be divided into two main categories: area-and feature-based methods.…”

“…Another issue is related to the differences in the image geometry acquired from these two devices [9,[11][12][13]15,16,18,19]. Thus, many co-registration approaches have been proposed (e.g., [9][10][11][12][13][14][15][16][17][18][19]), but, in general, these approaches might be divided into two main categories: area-and feature-based methods. Nowadays, the Structure from Motion (SfM) methods, which are mostly based on the feature-based approach, are used for the co-registration of spaceborne SAR and optical images [9,[11][12][13]15,16,18,19].…”

“…The presented technique was intensively tested, validated, and compared with the methods existing in the literature. Especially two approaches were investigated [9,17]. However, the method described in [17] is inefficient and fails due to the speckle noise on SAR images.…”