International Conference on Information Technology and Computer Science, 3rd (ITCS 2011)
79 Mean-Nearest-Neighbor: A More Pervasive, Fast and Adaptive Image Filtering Method
Download citation file:
- Ris (Zotero)
- Reference Manager
The concept of the Mean-Nearest-Neighbor was proposed basing on the analysis of the common filtering algorithm, and the appropriate filtering algorithm was designed. First of all, the mean of the remaining data was calculated after excluding some suspicious noise data by way of adapting to the size of the filtering windows. The second, three rules of calculating the mean and two programs of how to judge the noise were proposed, and subsequently nine filtering algorithms were also proposed basing on this. After experimental comparison with the mean filtering, the median filtering and some improved filtering algorithms, it is determined by this paper that the 7th filtering algorithm has a better filtering effect, which can adapt to different types of noise and enhance the universality and applicability.