PCA Based Image Enhancement in Wavelet Domain

This paper demonstrates a methodology of image enhancement that uses principle component analysis (PCA) in wavelet domain. PCA fully de-correlates the original data set so that the energy of the signal will concentrate on the small subset of PCA transformed dataset. The energy of random noise evenly spreads over the whole data set, we can easily distinguish signal from random noise over PCA domain. It consists of two stages: image enhancement by removing the random noise and further refinement of the first stage. The random noise is significantly reduced in the first stage; the Local Pixel Grouping (LPG) accuracy will be much improved in the second stage so that the final enhancement result is visually much better. The LPG-PCA enhance procedure is used to improve the image quality from first stage to second stage with edge preservation. The wavelet thresholding methods used for removing random noise has been researched extensively due to its effectiveness and simplicity. However, not much has been done to make the threshold values adaptive to the spatially changing statistics of images. Such adaptivity can improve the wavelet thresholding performance because it allows additional local information of the image (such as the identification of smooth or edge regions) to be incorporated into the algorithm. We compare this two-stage process with traditional principal component analysis and find that the results of the new structure are closer to the structure of traditional quality of image, its purity and descriptors than traditional principal component analysis. Keywords—Wavelet, Wavelet Transform (WT), Local Pixel Grouping (LPG), Principal Components Analysis.