Image Enhancement Using Discrete Wavelet Transform based on Color Space Conversion Technique

The image enhancement technique is most important technique to improve the quality of the image and visual appearance and to provide better representation for automated digital image processing. The major objective of the paper is to provide the novel algorithm to eliminate the Gaussian noise content and to improve the quality of the image. In this paper, the discrete wavelet transform (DWT) is proposed for smoothening and sharpening to improve the quality of the image. The main aim is to derive the wavelet transform coefficients in the new execution basis, the noise content can be removed from the image. The color image is considered for de-noising purpose to remove the noise content by applying DWT decomposition by converting the RGB (Red, Green and Blue) significance of each pixel of the original noise image to HSV (Hue, Saturation and Value) by giving comprehensive evaluation of proposed algorithm. The Gaussian noise is taken to the account and the proposed DWT algorithm is used to eliminate the Gaussian noise. The high frequency band is sharpened and low frequency sub-band is smoothened by applying non-linear filter. After converting RGB to HSV color space, the saturation band is enhanced by applying adaptive histogram equalization method. The V band is improved by applying illumination improvement and then all three HSV has to be converted back to RGB to reconstruct the original image with noise removal and high quality digital image.