Performance Analysis of DWT Families

Abstract

Discrete wavelet transform (DWT) is a technique which is used to acquire wavelets from image pixels. It transforms a discrete time signal to discrete wavelet representation. The DWT has shown great performance in digital image compression, demodulation, demultiplexing and denoising applications. In this paper, a comparison among different wavelet transforms is made. These wavelets include; Discrete Meyer, Biorthogonal, Fejer-korovkin, Daubechies, Coiflets, Symlets and Reverse Biorthogonal wavelets. Each wavelet is implemented for upto six levels of wavelet decomposition. The reconstructed images are then compared with the original images and their performance is evaluated with the help of Peak Signal-to-Noise Ratio (PSNR) and Mean Square Error (MSE). It has been observed in all wavelets that as we increase the decomposition level, MSE increases and PSNR decreases. In this paper, a comparison has been made between the different wavelet families, keeping the filter length fairly same in all the wavelets. This paper will help the new researchers of this field to choose the suitable wavelet transform for image processing applications.