A multiresolution approach for image morphing

A new multiresolution approach for image morphing is proposed in this research. By exploring the fact that the human visual system recognizes primarily global features of objects in a fast changing image sequence, we use mainly the low frequency information of images in morphing. Besides, we consider the morphing process in the wavelet domain rather than in the spatial domain and, as a consequence, both spatial and frequency characteristics of the initial and final images in a range of resolution scales can be effectively used. Experimental results of morphing of two face images are given 20 illustrate the performance of the proposed method.<>