Abstract:

In this paper Matlab-based application-GUI "Wavdec" is presented, serving as educational tool for different signal processing courses. Wavdec analyses an image into different wavelet subbands, using three different wavelet decomposition/reconstruction methods: Discrete Wavelet Transform (DWT), Quincunx transform and Steerable Pyramid Wavelet Transform (SPWT). Different levels and wavelet types for each method can be chosen. Also, Gaussian noise with manually defined mean and variance can be added to the original image. Hard thresholding, which either keeps or removes values of coefficients, for each subband can be defined arbitrarily. Thus, Wavdec can be used in different scenarios, for example to demonstrate scalable image compression or noise removal. Finally, Wavdec application automatically calculates image quality between original and reconstructed image using three image quality measures as well as their compression ratio. Finally, the effects of using the described original developed tool on education of electrical engineering students have been disclosed.

Keywords:

Educational tool, wavelet methods, Wavdec, signal processing, image quality.