SAR Image Matching Based on Speeded Up Robust Feature

Abstract

Speeded-Up Robust Features (SURF) is a novel scale-invariant and rotation-invariant feature. It is perfect in its high computation speed and robustness.In this paper, we apply SURF in SAR image matching accord to its characteristic, and then acquire its invariant feature for matching in an addition of no any pre-processing. In the process of image matching, we use the nearest neighbor rule for initial matching, whereafter, remove the wrong points of the matches through RANSAC. All this method was called R-SURF (RANSAC-SURF).In this method, the threshold range of the nearest neighbor rule has been obtained with our experiment .Experimental results indicated that the threshold interval was [0. 6～0.7], and the threshold that you choose in this interval, increased little matching time, but got more than 95% correct matching rate. We used three different types of SAR images in experiments which are in order to put to the proof that SURF is more robust in scale change, rotation change and noise.

Optimal texture image reconstruction method for improvement of SAR image matching

“…In this step, the SURF matching algorithm is applied to optimal texture images. SURF and SIFT methods are a common method for SAR image registration [1, 7, 31–33]. The main reason for using SURF instead of SIFT is the speed of the matching process [34].…”

Vision-based Runway Detection and Landing for Unmanned Aerial Vehicle Enhanced Autonomy

“…SURF is more efficient with respect to performance matching speed in comparison to SIFT [33]. In the area of matching images, therefore, Liu [41], Verma [42] and Vardhan [43] used SURF and verified that this approach has high performance and robustness. In regard to the performance of distorted image matching, Karami [44] compares the performance of SURF and SIFT, and the author denotes that SIFT has a higher matching rate in most rotation angles than SURF.…”

On Feature-Based SAR Image Registration: Appropriate Feature and Retrieval Algorithm

“…RANSAC [30] has been widely used in feature-based SAR image registrations for parameter retrieval [15,16,26,27]. Unlike LS which uses all the available data to estimate parameters, RANSAC conducts the estimation using a few-to-many strategy or a local-to-global strategy.…”

“…SURF has been demonstrated to outperform SIFT on speed, repeatability, distinctiveness, and robustness [8]. It has been used for multispectral satellite image registration [24], seabed recognition based on sonar images [25], and SAR image registration [26][27][28][29].…”

“…Therefore, the robust retrieval algorithms which are insensitive to outliers are needed. In many existing literatures on feature-based SAR image registration [15,16,26,27], the random sample consensus (RANSAC) [30] has been widely used and recommended for warp function retrieval. RANSAC conducts the estimation by randomly sampling a minimal sampling set (MSS) to achieve an estimation of the warping, and the entire datasets are then checked on the estimation for a consensus set (CS) of correspondences.…”