International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 185 - Number 30 |
Year of Publication: 2023 |
Authors: K. Sudhakar, R. Priyanka, V.Mani Kumar, A.V. Ramani, N. Aparna, M.Manju Sarma |
10.5120/ijca2023923059 |
K. Sudhakar, R. Priyanka, V.Mani Kumar, A.V. Ramani, N. Aparna, M.Manju Sarma . Automatic Image Registration Through-Principal Axes Registration. International Journal of Computer Applications. 185, 30 ( Aug 2023), 30-35. DOI=10.5120/ijca2023923059
Image registration aims to find one to one correspondence between points that are presented in images which are needed to be registered. Also Image registration is the process of spatially aligning two or more images of same location taken at different timings or from different view-points by transforming into one coordinate system. In this paper, a new automatic satellite image registration is proposed based on principal axes registration method. The proposed algorithm uses Scale-invariant feature transform (SIFT) to detect point features from high resolution satellite images and then for removing mismatches (outliers) it uses Random sample consensus (RANSAC) algorithm. Finally, principal axes registration method used to register two images which share common visual information and that are assumed to be differing from each other. The principal axes registration algorithm is based on affine transformation to register two images which are transformed by translation, rotation and scaling. For validation The Root Mean Square Error (RMSE) is used for similarity measure. Experiment results shows that the proposed algorithm can register high resolution satellite images (Cartosat-1, Landsat, Resourcesat-1) which are captured at different timings. The proposed algorithm has been implemented in MATLAB.