Image denoising based on Bidimensional Empirical Mode Decomposition

Abstract

Through the analysis of principle and process of image signal denoising, a kind of image denoising algorithm based on Bidimensional Empirical Mode Decomposition is proposed. This paper has improved the traditional Bidimensional Empirical Mode Decomposition method. Bidimensional Empirical Mode Decomposition method is used to decompose the image signal and selective denoising is done to decomposition result by applying self-adaptive median filtering. Denoising result can fully retain the non-stationary feature which is inherent in image signal and it also has the characteristics of strong self-adaption, flexibility and effectiveness. Its computation speed and computational accuracy are greatly increased. It is proved by experiment that when processing noising image, this method not only greatly reduces the noise, but also retains the detail information like the edge of original image well. Keywords-bidimensional separable empirical mode decomposition; denoising; adaptive median filteringI.

Moving Average Filter untuk Memisahkan Efek Dangkal Anomali Gravitasi Time Lapse

“…Pemisahan anomali bisa dilakukan dengan metode pemisahan spektrum (Xu et al, 2009). Penelitian tentang pemisahan efek dangkal dengan moving average filter pernah dilakukan oleh peneliti terdahulu (Pei, et al, 2012;Shandini & Tadjou, 2012;Indriana et al, 2018) akan tetapi tidak dilakukan pada gravitasi time lapse. Pada penelitian ini pemisahan anomali dengan moving average filter dilakukan pada data gravitasi 4D (time lapse).…”

Self-adaptive image denoising based on bidimensional empirical mode decomposition (BEMD)

“…Image denoising method based on BEMD has been used by some researchers. For example, Pei et al used the adaptive median filtering to remove the noise of each 2D-IMF obtained by BEMD [22]. In this work, the threshold was computed using the formula ( )…”

“…The filter width and shape could be varied from one 2D-IMF to the next. However, for users who are not expert in the images for denoising, it is difficult to determine the values of parameters such as N in [22] and the filter width and shape in [23]. Therefore, the selfadaptive image denoising based on BEMD could be very useful.…”