Adaptive Regularized Level Set Method for Weak Boundary Object Segmentation

Figure 4

Segmentation results on images with and without (Gaussian) noise. Row 1: original and noise images ((a) and (d) original images; (b) noisy image (zero mean, 0.04 variance); (c) noisy image (zero mean, 0.1 variance); (e) noisy images (zero mean, 0.01 variance); (f) noisy image (zero mean, 0.04 variance)). Row 2: results of the RSF model [8]. Row 3: results of the LIF model [18]. Row 4: results of our model (left to right: 𝜈 = 0.5, 0.28, 0.2, 0.4, 0.25, 0.21).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

(q)

(r)

(s)

(t)

(u)

(v)

(w)

(x)