A new method for designing an efficient switching median filter using VLSI architecture to remove salt and pepper noise

Any imaging system's images have dynamic intensity value variations, rapid light shifts, and poor contrast. Such visuals are empty of useful information and difficult to comprehend visually. Filtering methods to eliminate noise, improve contrast, and detect edges are used to retrieve secondary information from such photographs. Median filtering is one of the nonlinear ways that remove the range of isolated noise like salt and pepper noise while preserving the edge information of the image. However, median filtering fails to remove the noise when the image is affected by too strong impulsive noise. The switching median filtering technique can be used to eliminate high-intensity impulse noise. This research presents a VLSI design for a novel switching-based median filter to reduce high-density salt and pepper noise in digital images. The absolute difference between the center pixel and the array median obtained from a 3 x 3 sliding window is compared to a predefined threshold value to determine whether a pixel is noisy or not. During the filtering stage, the noisy pixels in the 3 x 3 filtering window are replaced by the median of noise-free pixels. If a pixel value is damaged, it is replaced with the median of the following window. The true pixel intensity value is kept if the pixel is not destroyed. We can tell if a pixel is corrupted or not by using a threshold detector. We may either add salt and pepper noise to an image directly or create it ourselves. This is a step in the image preparation process. The normal image is uploaded and converted it to a salt and pepper noise image using MATLAB's built-in capabilities. We used three tools in our project. We used MATLAB to preprocess the picture and plot data to reconstruct the original image, then ModelSim to produce the median filter, and XILINX to compare the existing bubble sort to the recommended three-cell sorter. The proposed technique outperforms standard median-based filters in simulations and is particularly effective in circumstances where pictures are badly damaged.