Title :
Nearest-neighbor image model
Author :
Sandryhaila, Aliaksei ; Moura, Jose M. F.
Author_Institution :
Electr. & Comput. Eng., Carnegie Mellon Univ., Pittsburgh, PA, USA
fDate :
Sept. 30 2012-Oct. 3 2012
Abstract :
We propose a novel, adaptive model for image representation. The model places image pixels on the nodes of a two-dimensional nearest-neighbor graph. Edge weights for the graph depend on the image of interest, and can be determined by solving a corresponding least-squares problem. The proposed model is shown to provide an efficient image representation well-suited for image compression.
Keywords :
graph theory; image coding; image representation; least squares approximations; 2D nearest neighbor graph; adaptive nearest neighbor image model; edge weight; image compression; image pixels; image representation; least square problem; Adaptation models; Discrete cosine transforms; Discrete wavelet transforms; Image coding; Image representation; Polynomials; Image representation; compression; nearest-neighbor graph; orthogonal polynomials; orthogonal transform;
Conference_Titel :
Image Processing (ICIP), 2012 19th IEEE International Conference on
Conference_Location :
Orlando, FL
Print_ISBN :
978-1-4673-2534-9
Electronic_ISBN :
1522-4880
DOI :
10.1109/ICIP.2012.6467411
