Figure 8
From: A photoacoustic image reconstruction method using total variation and nonconvex optimization
The different image objectives reconstructed by TV-GD and TV-Lp algorithms. The first to sixth rows refer to the original phantom (a, g, m), the TV-GD reconstructed images (b, h, n), the TV-Lp reconstructed images with parameter settings as (α = 1 × 10-2, β = 5 × 10-3) (c, i, o), (α = 1 × 10-2, β = 1 × 10-2) (d, j, p), (α = 5 × 10-2, β = 5 × 10-3) (e, k, q) and (α = 5 × 10-3, β = 5 × 10-3) (f, l, r), respectively.