A multiresolution spline with application to image mosaics

Abstract

We define a multiresolution spline technique for combining two or more images into a larger image mosaic. In this procedure, the images to be splined are first decomposed into a set of band-pass filtered component images. Next, the component images in each spatial frequency hand are assembled into a corresponding bandpass mosaic. In this step, component images are joined using a weighted average within a transition zone which is proportional in size to the wave lengths represented in the band. Finally, these band-pass mosaic images are summed to obtain the desired image mosaic. In this way, the spline is matched to the scale of features within the images themselves. When coarse features occur near borders, these are blended gradually over a relatively large distance without blurring or otherwise degrading finer image details in the neighborhood of th e border.

Curiosity’s Mars Hand Lens Imager (MAHLI) Investigation

“…The blending option combines the images using multi-resolution spline-based image blending (based on Burt and Adelson 1983). The few focus merge products produced with the testbed-MAHLI on the testbed rover on Earth (in the absence of Mars winds) using all of these options (merge-only, merge + registration, merge + blending, or merge + registration + blending) show no difference in the quality or resolution of the products (mainly because the camera did not move during focus stack acquisition).…”

Model-based human shape reconstruction from multiple views

“…Techniques for blending make use of factors such as the relative orientation of the surface with respect to a camera view and the relative distance to the edge of the texture to define the weighting term [52]. Here a multiple resolution blending technique is used as introduced by [53]. A multiple resolution approach ensures that the extent of texture blending corresponds to the spatial frequency of the features in the texture image, preserving the higher frequency detail in the resampled appearance.…”

Segmentation of elastographic images using a coarse-to-fine active contour model

“…2a. To accomplish the multiresolution approach, a Gaussian pyramid (Burt and Adelson 1983) was constructed by hierarchically convolving the original image with a spatial Gaussian kernel, smoothing with Gaussians (low pass filters) and resampling at each level. The pixel value in a level was the weighted average of pixel values in the next lower level.…”

“…The proposed algorithm is based on a coarse-to-fine transformation (Gaussian pyramid) (Burt and Adelson 1983) for active contour initialization and a gradient vector flow active contour model for deformable contour optimization, with the help of prior knowledge of the geometry of general thermal coagulations.…”