Title: Image denoising using fast non-local means filter and multi-thresholding with harmony search algorithm for WSN

Authors: H. Rekha; P. Samundiswary

Addresses: Department of Electronics Engineering, School of Engineering and Technology, Pondicherry University, Pondicherry, India ' Department of Electronics Engineering, School of Engineering and Technology, Pondicherry University, Pondicherry, India

Abstract: Several image denoising algorithms are developed so far to obtain a better denoised sensor images. But they fail to preserve the edges of the images. In order to overcome this, an attempt has been made in this paper by incorporating the various filters like fast non-local means filter (FNLMF) and high boost filter with the existing wavelet thresholding-based denoising method. However, the denoised output and the computation time are affected by the wavelet properties significantly. Hence, this paper concentrates on histogram-based multi-thresholding (HMT) as a major part to denoise the image. Here, the corrupted image is first denoised by applying FNLMF filtering section. Then the edges and minimal details of the denoised output are enhanced by utilising the HMT with harmony search algorithm (HSA)-based optimisation technique. From the simulation results, the proposed denoising method can be adoptable for WSN.

Keywords: bilateral filter; entropy; fast non-local means filter; histogram; image denoising; multi-thresholding; high boost filter; HBF; optimisation; harmony search algorithm; wavelet thresholding; wireless sensor network; WSN.

DOI: 10.1504/IJAIP.2023.128076

International Journal of Advanced Intelligence Paradigms, 2023 Vol.24 No.1/2, pp.92 - 109

Received: 24 Jan 2017
Accepted: 19 Dec 2017
Published online: 05 Jan 2023 *