30 October 2009 Image fusion via multi-orientation gradient mode of image block and non-down-sampled B-spline wavelet transform
Author Affiliations +
Proceedings Volume 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis; 74954B (2009) https://doi.org/10.1117/12.831313
Event: Sixth International Symposium on Multispectral Image Processing and Pattern Recognition, 2009, Yichang, China
Abstract
A method for image fusion based on multi-orientation gradient mode and non-down-sampled B-spline Wavelet Transform is proposed, by using block partition of the image fusion method of thinking. Multi-directional gradient filters are designed by directional derivative. High-frequency image blocks are distilled by non-down-sampled B-spline wavelet transform, and are filtered by multi-directional gradient filters. Compared with the total value of the gradient of corresponding high-frequency image block is in order to pick out a clear image of the block .Finally, the final fusion image can be achieved by reconstructing clear picture block and application of the consistency constraints. This method doesn't need wavelet inverse transform to reconstruct image, reserve the original images to achieve high-precision integration. Experiments show that the method for multi-focus image fusion has good results and good stability.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Zhen-kun Lu, Zhen-ming Yu, "Image fusion via multi-orientation gradient mode of image block and non-down-sampled B-spline wavelet transform", Proc. SPIE 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis, 74954B (30 October 2009); doi: 10.1117/12.831313; https://doi.org/10.1117/12.831313
PROCEEDINGS
8 PAGES


SHARE
Back to Top