The design of rotation-symmetric Gaussian low-pass filter (RSGLPF) and its applications.
Tongnan Xia,Ming Liu,Jie Sun,Enruo Huang,Shaolin Liang,Yaojie Sun
DOI: https://doi.org/10.1145/3614008.3614060
2023-01-01
Abstract:In this study, a rotation-symmetric Gaussian low-pass filter (RSGLPF) is designed, and two application methods of the filer are provided. The results of the two methods are different, and researchers and engineers can choose which method to use according to their own needs. In data related research, noise reduction is one of the key steps in most data processing. However, for many non-data processing researchers, it is crucial to choose a simple and effective noise reduction algorithm. Additionally, traditional Gaussian filters could cause some degree of detail damage in the image, we have provided a method to effectively solve this problem. In the present research, there are many low-pass filters. Some of them have simple applications, but the filtering effect may not be very obvious. Others have complex application processes or require high parameter settings. Our research results have both a significant filtering effect and an easy-to-use application process. Even non-experts in this field can apply by following the steps. Moreover, the traditional Gaussian filter will cause some detail damage to images in the process of denoising. We have provided a method that can effectively reduce the detail damage of the original image. Gaussian filter, as a linear smoothing filter, is suitable for eliminating Gaussian noise and has been widely used in image processing, signal processing and other fields.[1, 3] Therefore, our (RSGLPF) is improved on the original Gaussian filter, and can be better applied to the noise reduction process. The results can also show that the denoising effect is indeed more obvious. The results show that the image is smoother, the boundary is more blurred, and the high-frequency signal of the waveform is greatly reduced. The research results provided a practical algorithm model for denoising in signal processing.