ABSTRACT
Image Mosaicing is an active research area in computer vision and computer graphics. Image mosaicing is the process of combining two or more images of the same scene into one high resolution image which is called panoramic image mosaicing. Image mosaicing techniques can be categorized into two general approaches: direct and feature based techniques. Direct techniques compare all the pixel intensities of the images with each other, whereas feature based techniques aim to determine a relationship between the images through distinct features extracted from the processed images. Feature extraction is the process of extracting the invariant features from the images. Several methods were proposed earlier for the feature extraction like Harris corner detector and scale invariant feature transform (SIFT), which are difficult and time consuming for real time applications. The wavelet feature extraction is not effective because of decomposing the image into wavelet subspaces at various scales and extracting key point by decomposing low frequency coefficients to get high frequency coefficients. And also wavelets will give an orientation variant feature which is not suitable for much application like image mosaicing. To achieve a better feature extraction a novel method is introduced and implemented by using complex wavelet packets. Here the image decomposed into wavelet subspaces at various scales and extracting key point by decomposing high frequency coefficients to get high frequency coefficients. This makes feature extraction very simple and more effective at different scales to get strong key point. Implementation results show that the feature extraction by using the proposed method is more reliable for real time applications when compared with existing methods.
