Unconstrained automatic image matching using multiresolutional critical-point filters

Abstract

This paper proposes a novel method for matching images. The results can be used for a variety of applications: fully automatic morphing, object recognition, stereo photogrammetry, and volume rendering. Optimal mappings between the given images are computed automatically using multiresolutional nonlinear filters that extract the critical points of the images of each resolution. Parameters are set completely automatically by dynamical computation analogous to human visual systems. No prior knowledge about the objects is necessary. The matching results can be used to generate intermediate views when given two different views of objects. When used for morphing, our method automatically transforms the given images. There is no need for manually specifying the correspondence between the two images. When used for volume rendering, our method reconstructs the intermediate images between cross-sections accurately, even when the distance between them is long and the cross-sections vary widely in shape. A large number of experiments has been carried out to show the usefulness and capability of our method.

Image Fusion using Gaussian Mixture Models

“…[1, 6,14,17]. Figure 1 contains a set of images with varying degrees of masking and the final fused (9)- (10); tests performed on the standard image database with noise level σ =25; the average increase in PSNR is 0.42. result using the proposed model.…”

Video metamorphosis using dense flow fields

“…On the other hand, the critical point filters adopted in Ref. [8] can effectively preserve important features such as local maxima, local minima and saddle points in lower resolution images. Inspired by the advantage offered by the critical point filters, we build a hybrid image pyramid for each input image.…”

“…Inspired by image matching in computer vision 3,8 , we designed a new algorithm for generating dense mappings in an optimization framework.…”

Palmprint verification using hierarchical decomposition

“…They can be categorized into two groups: linear filters and nonlinear filters [30]. The characteristic of MRFs is that they will directly discard the pixels while doing down sampling.…”

“…The information at different resolutions in an image generally represents different physical structures in the image. In the literature a number of applications of multi-resolution representations in image analysis have been discussed [14,[30][31][32].…”