An Improved Architecture for Removal of Random Valued Impulse Noise To Obtain Quality Images
In a real time image processing systems, images are habitually corrupted by impulse noise during the process of image acquisition and communication. Hence there is a need for an efficient and manipulator friendly impulse noise removal technique. In this paper an efficient noise eradication scheme for the removal of random-valued impulse noise in images and its low-depletion VLSI architecture has been proposed. The architecture consists of two line buffers, register banks, impulse noise detector and an edge preserving filter. To detect the noisy pixels, a verdict investigation based impulse noise detector is used, and an edge-preserving filter is employed to reconstruct the intensity values of noisy pixels. To enhance the effects of noise removal, an adaptive technology is used. The proposed storage hardware is multi line buffer rather than a full framed memory and the proposed algorithm involves only a fixed size window instead of variable window size. These features greatly reduce the memory requirement as well as computational complexity. The proposed edge preserving algorithm provides better visual quality for noise free image. The image quality assessment parameters such as SNR, PSNR, Mean Squared Error and Structural Similarity are computed for the reconstructed noise free image.