A physics-inspired model of image structure representation by deformable elastic grid
Xiaodong Zhuang, Nikos E. Mastorakis · Computational intelligence · 2008
A novel physics-inspired model is proposed for image structure representation. The deformable elastic grid is defined on digital images. The attracting force between adjacent points is defined according to the gray-scale difference, which is the source of force causing the deformation of the elastic grid. The final shape of the grid after deformation can represent image structure information, based on which a segmentation method is proposed for digital images. Experimental results indicate the effectiveness of the proposed method.