A comparative study of decomposition level using entropy and wiener function in wavelet transform
The fundamental requirement of applying wavelet transform on an image is to select the wavelet family and decomposition level. The selection of the wavelet family depends on the nature of the application and properties of the image/ signal. The selection of the decomposition level depends on the availability of fine details at the next level. Earlier the wavelet decomposition level was selected manually but in the recent times, there are various parameters introduced using which the level of decomposition is decided. Out of them entropy and wiener cost function are the efficient metrics for this purpose as they analyses the texture and noise level of the image respectively. This paper shows the comparative study of deciding level of wavelet decomposition using entropy and wiener cost function. The entropy metric decomposes the image on the basis of texture analysis and wiener function decomposes the image on the basis of the power of the additive noise present in the image.