Texture analysis using local fractal dimension of complex networks GONCALVES, D. N. ; SILVA, L. A. ; ARAUJO, R. F. S. ; MACHADO, B. B. ; GONÇALVES, W. N. XI Workshop de Visão Computacional, 2015. p. 236-241
The texture analysis is one of the most important research areas in computer vision. Currently, complex network has emerged as an approach for representing images due to its flexibility for modeling several problems. Generally, the application of the complex network theory involves two steps: representing the structure of interest into a network and extracting measurements from it. Due to the large number of measurements, decide the appropriate measurements for a particular problem is a challenge. This paper proposes the fractal dimension of complex networks as a measurement for extracting relevant features from texture images. For this, an image is modeled as a network where each pixel is mapped into a vertex. By varying the radius of connection between vertices, the proposed method estimates the local fractal dimension for each vertex to form a histogram of dimensions. Results on two image databases show the effectiveness of the proposed method against tradditional texture methods.