A Novel Technique of Image Mosaicing based on Discrete Wavelet Transform and its Performance Evaluation

Abstract

of camera or other imaging sensor regarding the area coverage in a single shot at a fixed resolution has coined the term Image mosaicing. It is an application of image registration which aligns images taken in different camera coordinate system (i.e. different orientation of camera). In this paper, a new technique of image mosaicing has been proposed in which Discrete-wavelet Transform(DWT) based feature detector and scale-invariant feature transform(SIFT) based feature descriptor has been used which makes this technique robust to image zooming, image rotation, affine distortion and illumination change due to varying light condition during image acquisition. Result shows the improved performance with respect to number of feature points detected and computation time. Also, quantitative evaluation parameter like structural similarity index (SSIM), normalized absolute error (NAE), etc. proves the better performance of this technique compared to other techniques till date.