Redes complexas na extração de características em imagens de peixes

Abstract

This paper presents an approach to image analysis using the techniques of Complex Networks for obtaining similarities between images. Initially, the features are extracted from the edge of images, which will be understood computationally as arrays. Through these matrices are generated adjacency matrices, which will be used to check the connections of the vertices (pixels) of images. Extracted from the adjacency matrix will be performed to extract the degree of the vertices to a threshold T and the extraction of the average degree of the vertices. In the extraction of the degree of the vertices each threshold is applied in the calculation and returns an array that will be analyzed with other matrices extracted by checking the likelihood of isomorphism.

Keywords

• Feature extraction
• Complex networks
• Graph theory
• Degree of vertices
• Adjacency matrix