Quaternion-Based Color Image Smoothing Using a Spatially Varying Kernel

Authors:
Özlem N. Subakan;Baba C. Vemuri
Affiliations:
Department of Computer and Information Science and Engineering, University of Florida, Gainesville, USA 32611;Department of Computer and Information Science and Engineering, University of Florida, Gainesville, USA 32611
Venue:
EMMCVPR '09 Proceedings of the 7th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition
Year:
2009

Citing 11
Cited 0

Animating rotation with quaternion curves

SIGGRAPH '85 Proceedings of the 12th annual conference on Computer graphics and interactive techniques
Images as Embedded Maps and Minimal Surfaces: Movies, Color, Texture, and Volumetric Medical Images

International Journal of Computer Vision - Special issue on computer vision research at the Technion
Multi-Dimensional Signal Processin Using an Algebraically Extended Signal Representation

AFPAC '97 Proceedings of the International Workshop on Algebraic Frames for the Perception-Action Cycle
Das Konzept einer zweidimensionalen Phase unter Verwendung einer algebraisch erweiterten Signalrepräsentation

Mustererkennung 1997, 19. DAGM-Symposium
Fast Anisotropic Smoothing of Multi-Valued Images using Curvature-Preserving PDE's

International Journal of Computer Vision
Multi-fiber reconstruction from diffusion MRI using mixture of wisharts and sparse deconvolution

IPMI'07 Proceedings of the 20th international conference on Information processing in medical imaging
Hypercomplex correlation techniques for vector images

IEEE Transactions on Signal Processing
Full-diversity, high-rate space-time block codes from division algebras

IEEE Transactions on Information Theory
Color TV: total variation methods for restoration of vector-valued images

IEEE Transactions on Image Processing
Color image enhancement via chromaticity diffusion

IEEE Transactions on Image Processing
Hypercomplex Fourier Transforms of Color Images

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Addressing the issue of feature/detail preserving color image smoothing, we propose a novel unified approach based on a quaternion framework. The main idea is to holistically extract the local orientation information at each lattice point, and then to incorporate it into the smoothing process. We introduce a new Quaternion Gabor Filter to derive the local orientation information in color images. This derived orientation information is modeled using a continuous mixture of appropriate exponential basis functions. We solve the continuous mixture integral in analytic form, and develop a spatially varying kernel which respects to the local geometry at each lattice point in a color image. Superior performance of our smoothing framework is demonstrated via comparison to competing state-of-the-art algorithms in literature.