Research on Salt-and-Pepper Noise Removal in Images

This research program for salt-and-pepper noise removal in images is compiled under MATLAB environment and primarily employs the median filtering method for noise elimination. The program enables users to effectively address salt-and-pepper noise issues in images, thereby enhancing image quality and clarity. The median filtering algorithm operates by sorting pixel values within a defined neighborhood window (typically 3x3 or 5x5 kernels) and selecting the median value as the new pixel value, effectively eliminating impulse noise while preserving edge details. Through this implementation, users can conveniently adjust median filtering parameters such as kernel size and iteration count to adapt to various image types and noise conditions. The core MATLAB functions involved include medfilt2() for 2D median filtering and custom implementations for handling boundary conditions. This program serves as a practical tool with broad applications in digital image processing, particularly in preprocessing stages for computer vision tasks.