Super-Resolution Reconstruction of Video Sequences by GST-Driven Adaptive Filtering

Abstract:

Article Preview

A GST-driven spatially adaptive filter is developed in this paper based on the framework of non-local means (NLM) avoiding explicit motion estimation. Gradient Structure Tensor (GST) is introduced to express the underlying local image structural patterns, which drives the window function to yield adaptive scale and shape fitting for the local structure. This leads to patches with more similar gray-level and local structures being gathered for super-resolution estimation of image. Results on several test video sequences show that the proposed method is effective in providing super-resolution on general sequences and achieves improvement of performance on the compared method.

Info:

Periodical:

Advanced Materials Research (Volumes 255-260)

Edited by:

Jingying Zhao

Pages:

2145-2149

DOI:

10.4028/www.scientific.net/AMR.255-260.2145

Citation:

L. Guo and L. Z. Lian, "Super-Resolution Reconstruction of Video Sequences by GST-Driven Adaptive Filtering", Advanced Materials Research, Vols. 255-260, pp. 2145-2149, 2011

Online since:

May 2011

Authors:

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.