Removal of impulse noise using decision tree based denoising method

Images will be corrupted by impulse noise in image acquisition systems. Denoising can enrich the visual quality. Denoising plays an important role in medical imaging and scans where very small amount of noise present may be misinterpreted as an abnormality in human health system. In this paper, a denoising scheme for the removal of impulse noise with random value is worked out. The corrupted pixels are recognized by a noise detector. The intensity values of noisy pixels are reconstructed by a filter. The filter may be of edge preserving one. In this work, the corrupted pixels are reconstructed and the remaining pixels are kept untouched. VLSI implementation of the proposed work is carried out using MODEL-SIM ALTERA and the results are compared for peak signal to noise ratio (PSNR) in various compositions of noise added.