A magneto-statics inspired transform for structure representation and analysis of digital images
Xiaodong Zhuang, Nikos E. Mastorakis · 2009
Abstract:- Physical-field inspired methodology has become a new branch in image processing techniques. In this paper, a novel image transform is proposed imitating the source reverse of magneto-static field. The image is taken as a vertical magnetic field, and its curl is estimated as the virtual source of the field for image structure representation and analysis. The restoration from the virtual source to the image is also investigated, based on which a method of image estimation and restoration from its gradient field is proposed. The experimental results indicate that the proposed curl source reverse provides effective image structure representation, which can be exploited in further image processing tasks. Key-Words:- Image transform, curl source reverse, magneto-static, image structure, image restoration 1