Performance Evaluation of Image Enhancement Techniques in Spatial and Wavelet Domains

Abstract

Denoising is one of the important tasks in image processing. Despite the significant research conducted on this topic, the development of efficient denoising methods is still a compelling challenge.

In this paper, spatial domain methods and Wavelet Domain Methods of image denoising have been evaluated. The medical ultrasound images suffer from speckle noise which is multiplicative in nature and more difficult to remove than additive noise. In the spatial filter methods Median Filter and Wiener Filter are implemented. These methods are based on the simple formulas that are proposed by different authors. In Wavelet Methods Visu Shrink, Neigh shrink and Bayes Shrink are implemented. The basic idea of wavelet methods is to denoise the image by applying wavelet transform to the noisy image, then thresholding the detailed wavelet coefficient and inverse transforming the set of thresholded coefficient to obtain the denoised image.

The comparison of all filters methods is done using various Quality Metrics like Peak Signal-to-Noise Ratio (PSNR), Bit Error Rate (BER), Mean Square Error, etc.

The filters methods implemented in MATLAB 7.10.0.499(R2010a).