An Efficient VLSI Architecture For Removal Of Impulse Noise In The Image Using Edge Preserving Filter

  • Shreejay Darne, S.P.Dhanure, D.S.Nikam, Sumit Nagpure, Rutuja Yelmate


 Impulse noise is a common cause of image corruption during image processing and transmission procedures. To achieve low-complexity VLSI architecture, we propose An Efficient VLSI Architecture for Removal of Impulse Noise in Images Using Edge Preserving Filters in this paper. To detect the noisy pixels, we use a decision-tree-based impulse noise detector, as well as an edge-preserving filter. To recreate the intensity values of noisy pixels, an edge-preserving filter and a mathematical morphological filter were used, as well as an edge-preserving filter and a mathematical morphological filter. In addition, adaptive technology is used to boost the effects of removing random valued impulse noise. In terms of both quantitative assessment and visual efficiency, our proposed methodology outperforms the previous lower-complexity approaches. Furthermore, the results could be equivalent to those of higher-complexity methods. Using TSMC 0.18m technology, the VLSI architecture of our design achieves a processing rate of around 200 MHz the architecture only needs two-line memory buffers and has a low computational complexity. It has a low hardware cost and can be used in a variety of real-time applications.