Denoising of tube-type bottle image based on independent component analysis and nonsubsampled contourlet transform

In this paper a new image denoising algorithm is presented based on independent component analysis(ICA) and nonsubsampled contourlet transform(NSCT), taking full advantage of NSCT´s strong points of translation-invariant, multidirection-selectivity and ICA´s strong point of higher order statistical property, then a noisy image is denoised by maximum likelihood estimation of the noisy version of the ICA model. The simulation results have shown that the performance of the above method is superior both in signal to noise ratio(SNR) and edge preservation. This algorithm is suitable for defects monitoring systems in tube-type bottle.