hybrid methodology
1 article · search the full text for this term
-
Multispectral Image Denoising using Bi-Directional Recurrent Neural Network with DPCA Algorithm
Abstract: In practices, a Multispectral Images (MSI) image is always prone to corruption by various sources of noises while procuring the images. In this paper we implemented Decomposable Pixel Component Analysis (DPCA) algorithm with recurrent neural network (RNN) which effectively denoised the MSI images. The RNN enables shrinkage of non-reusable data points during DPCA execution and aid it to effectively transform spatial coordinate in a logical manner and reduces the time …
Published in Journal of Image Processing & Pattern Recognition Progress · Vol. 2, Issue 1, 2015 · pp. 25–30 Read article