Active contours method has been successfully applied to image segmentation even under low signal/noise ratio situations. These methods consist of total energy minimization, calculated by the sum of internal energy, which is a function of active contour geometry, and external energy which is function of pixel intensity of the image to be segmented. Traditionally, external energy is calculated by input image gradient. Among the methods of active contours, radial methods have lower computational complexity and have real time applications. In this work we propose the use of Hilbert Transform calculation for external energy in radial active contours. This external energy is named Hilbertian energy. Tests were performed synthetic images and show that Hilbertian energy is able to avoid the use of derivative and balloon force, performing segmentation properly.
Mais detalhes em:
Nenhum comentário:
Postar um comentário