An Improved Adaptive Median Filtering Algorithm Based on Star Map Denoising

Hancheng Cao,Naijun Shen,Qian Chen
2023-01-01
Abstract:In order to improve the filtering performance of salt and pepper noise in starry sky images, an improved adaptive median filtering algorithm is proposed. Firstly, the algorithm performs adaptive median filtering on the noise image. Instead of pixel median replacement, it obtains the set of suspected noise points. Then, it uses a neighbourhood differential noise template for secondary detection of the suspected noise points to obtain the real noise points. Finally, a weighted mean value is calculated based on the distance weights, and the noise points are replaced. The proposed algorithm is used to filter the star maps containing 20%–90% salt and pepper noise. The experimental results show that the denoising effect of this algorithm is better than other algorithms, while preserving the image details and edge information better.
What problem does this paper attempt to address?