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

Abstract

In this paper, the problem of Feature-Based Image Registration (FBIR) has been addressed using a novel hybrid feature detector. To prove the efficiency of the proposed hybrid feature detector, its performance is compared with that of the state-of-the-art feature detectors like BRISK, FAST, ORB, Harris, MinEigen and MSER. Three different image acquisition methods are choosen for comparison. They are (i) Rotation transformation (ii) Scene to model transformation and (iii) Scaling transformation. Taking into consideration the above three types of image acquisition methods, a FBIR method is proposed for different types of remote sensing images. This paper proposed a hybrid feature detection algorithm, that consumes less time to detect the feature key points and it is also rotation and scale invariant. In addition to this the proposed algorithm performs well in terms of match points and match rate. This paper also focuses on a comparative analysis of BRISK, FAST, Harris, MinEigen, MSER and ORB with the proposed algorithm in terms of time to detect feature keypoints. It can be observed from the results and tables that in the case of hybrid feature detectors,It takes less time to detect the feature key points.