W. Snyder et al., IMAGE RELAXATION - RESTORATION AND FEATURE-EXTRACTION, IEEE transactions on pattern analysis and machine intelligence, 17(6), 1995, pp. 620-624
The techniques of a posteriori image restoration and iterative image f
eature extraction are described and compared. Image feature extraction
methods known as Graduated Nonconvexity (GNC), Variable Conductance D
iffusion (VCD), Anisotropic Diffusion, and Biased Anisotropic Diffusio
n (BAD), which extract edges from noisy images, are compared with a re
storation/feature extraction method known as Mean Field Annealing (MFA
). Ail are shown to be performing the same basic operation: image rela
xation, This equivalence shows the relationship between energy minimiz
ation methods and spatial analysis methods and between their respectiv
e parameters of temperature and scale. As a result of the equivalence,
VCD is demonstrated to minimize a cost function, and that cost is spe
cified explicitly. Furthermore, operations over scale space are shown
to be a method of avoiding local minima.