Efficient Multisource Remote Sensing Image Matching Using Dominant Orientation of Gradient

Abstract

Image matching is the key step for image registration. Due to the existing nonlinear intensity differences between multisource images, their matching is still a challenging task. A fast matching approach based on dominant orientation of gradient (DOG) is proposed in this article, which is robust to nonlinear intensity variations. The DOG feature maps are constructed by extracting DOG feature of each pixel in the images in the first place. A template matching method is used to determine correspondences between images based on the feature representations. We define a similarity measurement, referred to as sum of cosine differences, which can be accelerated by fast Fourier transform. Subsequently, the subpixel accuracy can be achieved by fitting the similarity measurement using a quadratic polynomial modal. A new variable template matching (VTM) method has been developed to improve the matching performance. Experimental results confirm that the proposed matching approach is robust to nonlinear intensity differences and has time efficiency. The VTM method additionally improves the matching precision effectively. Index Terms-Dominant orientation of gradient (DOG), image matching, variable template matching (VTM).

Multimodal Remote Sensing Image Registration Methods and Advancements: A Survey

“…In DDGWO, a support vector machine (SVM) model was trained to predict HOGD instead of calculating it directly, which can lead to a significant reduction in computation time. Liang et al [64] proposed a fast-matching method based on dominant orientation of gradient (DOG), which constructed a feature map by extracting the DOG feature of each pixel in an image. The author defined a new similarity measure called the sum of cosine difference, which can be accelerated by the Fast Fourier Transform (FFT).…”

Fast Patchwise Nonlocal SAR Image Despeckling Using Joint Intensity and Structure Measures

“…To preserve the structure details, the structure similarity is measured in the paper. The gradient orientation has been proved to be able to measure the structure similarity between remote sensing images, even for multi-source remote sensing images [40]. Hence,we attempt to extract gradient orientation from an amplitude image as a feature map to be measured.…”

“…where o(x + t) and o(x) represent the gradient orientation at coordinates (x + t) and x, respectively. The distance is not exactly the same as the one with extra coefficient 2 in cos(•) proposed in [40], because the gradient direction here belongs to [0, 2π) instead of [0, π) for multi-source remote sensing images.…”

A Hybrid approach for Image Acquisition Methods based on Feature Based Image Registration (FBIR)

“…In another recent work, presented in [24] , the authors describe a comprehensive survey on different feature based image matching procedures also about feature detection and description methods. D. Liang et al [25], proposed an image matching algorithm known as "Dominant Orientation of Gradient (DOG)" and this technique is found to be robust to "nonlinear intensity variations". Very recently that is in the year 2022, X. Liu et al [26] proposed a novel local statistics based image registration scheme.…”