Preserving Edge from Gaussian Filter by Edge Template
Nattapong Jundang, Surachai Ongkittikul · 2018
An image filtering with the only Gaussian function is the smoothing technique. Although the result of the filtering can remove the noise, it also destroy the detail image. This paper presents the smoothing technique that uses the Gaussian function to locate the image edge in any position. By using the edge model template, the comparison has been make to adjust the proper kernel of Gaussian function. If the decision was made to be edge, the proper kernel was adjusted respect to the orientation of the edge that was apply. Otherwise, the normal filter was apply. The result of our approach will performed the same as Gaussian function but can preserve the edge in the image. The experiments will demonstrated the various of the parameters such as the kernel and σ. The results were shown that our approach can performed the significant in term of the preserving edge.