Professional Documents
Culture Documents
Poster CAIP 2013 - Inpainting
Poster CAIP 2013 - Inpainting
I
initialization i=0
=
P
(a)
(b)
C'
i=1
step 1
priority computation
patch reconstruction
Always pathological cases of reconstruction artefacts (cf. Fig. 1(b)) Appearance of seams between reconstruction patches
Solution 1. Artefact detection locations of locally bad reconstructions 2. Spatial patch blending reconstruction patches seams reduction 1. Artefact detection a. Reconstruction artefact locations ? i) Existence of sharp variations in I high I ii) Reconstruction patch locations U locally very dierent high div (U ) b. Break eld R(p): strength of artefacts, combination of hypothesis i) and ii) p , where is a normalization factor. c. Blending amplitude map
wb(p,r) I (p) | div(U )(p) |
R(p ) =
p ,
rE (p) = max
q rE
wb(q,r)
patches
Equation: compose an image J with a set p of reconstruction patches q centered at each q , from a neighbourhood of p J (p ) =
i
q p i w(q,p) q (pq )
+
q p
w(q,p)
with w(q, p) a gaussain weight based on a distance from q to p. Our method is already embedded inside a GMIC plugin for GIMP: http://gmic.sourceforge.net/gimp.shtml
http://www.greyc.fr/