Abstract
A new enhancement method for aviation images is proposed, which combines improved nonsubsampled contourlet transform (NSCT) with guided filtering together. The original image is decomposed into a low pass sub-band and several high pass sub-bands. To improve the color contrast ratio of the image and remove the noise, fuzzy set theory is adopted to process the low frequency coefficients. As for high pass sub-band, those coefficients are optimized and enhanced by nonlinear gain function first, in which the thresholds is adaptive with Bayes shrink. After transforming back from the NSCT domain, the reconstructed image is sent to guide filter for final enhancement of edge details. Experiments show that compared with other comparative methods, the proposed algorithm is not only better in visual effects, but also superior to performance in those objective evaluation criterions.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
