Monochrome Image Impulse Noise Removal Considering Line Structure

Impulse noise affects images during compression or transmission, reducing their quality. Considering the line structures of images during the restoration process of pixel values contaminated by impulse noise can enhance the image's quality. The Line Structure Filter (LSF) method not only eliminates impulse noise but also effectively restores line structures in images. Here, we introduce a dual filtering strategy based on LSF. This strategy utilizes two differently sized filters: the first detects line structures in horizontal and vertical directions; the second expands to four directions. Filtering window selection is based on regional characteristics. Combining different filtering windows more effectively preserves image line structures. Experimental results show that our algorithm outperforms existing methods in removing impulse noise and restoring line structures.