problemname TeddyStereo1
width 450
height 375
labels 60
labelscale 4
smoothexp 1
smoothmax 1
lambda 10
datacost teddy/datacost%03d%c.pgm
datascale 4000
smoothx teddy/smoothx%c.pgm
smoothy teddy/smoothy%c.pgm
smoothscale 20000

The parameters used for cost and smoothness images:

match_fn 1   # SAD
match_max 16    # truncated, i.e. cost = |min(diff, 16)|
match_interval 1 # Birchfield / Tomasi

opt_grad_thresh 10 # i.e. if gradient <= 10, then
opt_grad_penalty 3 # multiply smoothness cost by 3