30 October 2009 Research on regularization super resolution reconstruction algorithm based on ASTER image
Author Affiliations +
Proceedings Volume 7498, MIPPR 2009: Remote Sensing and GIS Data Processing and Other Applications; 749822 (2009); doi: 10.1117/12.833046
Event: Sixth International Symposium on Multispectral Image Processing and Pattern Recognition, 2009, Yichang, China
Abstract
Super resolution reconstruction is to produce one or a set of high resolution images from a sequence of low resolution images using the additional information among them. Traditional super resolution reconstruction algorithms are limited to their assumed data and noise model. The robust reconstruction algorithm which is not sensitive to model error has always been a hot research. We propose an alternate approach based on p-norm minimization and robust regularization with bilateral total variation (BTV). This method is robust to errors caused by motion and blur estimation. Hybrid steepest descent and limited storage quasi-Newton method is used to solve the cost function. Experiments are carried out in simulated images and ASTER multi-band thermal infrared images, experiment results indicate that the proposed method removes noises effectively and results in fine detail, sharp edge and rich content.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Lin Yao, Yingbao Yang, "Research on regularization super resolution reconstruction algorithm based on ASTER image", Proc. SPIE 7498, MIPPR 2009: Remote Sensing and GIS Data Processing and Other Applications, 749822 (30 October 2009); doi: 10.1117/12.833046; https://doi.org/10.1117/12.833046
PROCEEDINGS
8 PAGES


SHARE
KEYWORDS
Super resolution

Lawrencium

Image processing

Reconstruction algorithms

Image fusion

Data modeling

Infrared imaging

Back to Top