Discrete wavelet transform approach for image registration using adaptive polar transform

Image registration is an essential step in many image processing applications that involve multiple images for comparison, integration or analysis such as image fusion, image mosaics, image or scene change detection, and medical imaging. Image registration methods can be categorized into two major groups: the feature-based approach and the area-based approach. In area-based approach LPT was introduced for its scale and rotation invariant properties. LPT suffers from the non-uniform sampling and bias matching problem. Thus APT was introduced that requires less computational cost in the transformation process than that of the conventional LPT, while maintaining the robustness to the changes in scale and rotation using a projection transform. In order to reduce the computational load, feature points are selected, using which APT is applied to register the image. Gabor transform is one of the methods that can be used to find the feature points. The feature points generated so, are present at the edges of the image also. The presence of feature points along the edges cannot be used to register an image, as they are prone to variations when scaled or rotated. This problem has been avoided by first applying DWT to the image and then giving the low frequency component as the input to the Gabor transform to find the feature points.