Hybridization of DBA-DWT Algorithm for Enhancement and Restoration of Impulse Noise Image

Abstract

Data is transmitted as image or video. Later communication, the image found is frequently degraded with noise. The corrupted values in extreme ranges are called as salt and pepper noise. The filtering is generally utilized for elimination of impulse noise or salt and pepper noise for noise free images and are fully recovered by minimum signal distortion and furthermore uncorrupted the images. Median filter is utilized to remove the salt and pepper noise from the image. Median works in low densities however not in higher densities in light of the fact that at higher densities the images are blurred and damages the image. To overcome this a novel Decision based Algorithm with Discrete wavelet transform method is proposed. It works well in higher density conditions then discrete wavelet transform has been useful to the result of the decision based algorithm processed by image enhancement technique. Firstly the restored Image has been converted to HSV color transformation that is Hue, Saturation and Value. Later, luminance intensity of low frequency component of value (V) component is enhanced. It is then followed by contrast enhancement technique, resulted in modified (V) component. The adapted (V) component is combined with Hue (H) and Saturation (S) and converted back into enhanced RGB image. Experimental results improve the Images which are captured at low-light and non-uniform lighting condition as in several standard Images as well as in real time Image.