How to show GMM manual computation with MATLAB

2 Ansichten (letzte 30 Tage)
Fitroh Amaluddin
Fitroh Amaluddin am 24 Sep. 2014
Bearbeitet: Fitroh Amaluddin am 5 Jan. 2015
how i can get this result from RGB color with gaussian mixture model(GMM) computation in MATLAB?
like this output example:
=======================
***** iteration 1 *****
R:90 G:150 B:40
-= CHECK AND UPDATE =-
check with GMM 1
-179.056942<90<-20.943058 --> not in range
update weight GMM 1 : 0.099000
check with GMM 2
-179.056942<90<-20.943058 --> not in range
update weight GMM 2 : 0.099000
check with GMM 3
-179.056942<90<-20.943058 --> not in range
update weight GMM 3 : 0.099000
check with GMM 4
-179.056942<90<-20.943058 --> not in range
update weight GMM 4 : 0.099000
check with GMM 5
-179.056942<90<-20.943058 --> not in range
make new model :
mean R:90 mean G:150 mena B:40
weight: 0.100000
varian: 100.000000
update weight GMM 5: 0.109000
valaue GPD GMM 5: 0.000001
update mean GMM 5:
R:90.000000 G:150.00000 B:40.000000
update variance GMM 5: 99.999936
-= NORMALIZE WeiGHt =-
before: after:
weight GMM 1: 0.099000 weight GMM 1: 0.196040
weight GMM 2: 0.099000 weight GMM 2: 0.196040
weight GMM 3: 0.099000 weight GMM 3: 0.196040
weight GMM 4: 0.099000 weight GMM 4: 0.196040
weight GMM 5: 0.099000 weight GMM 5: 0.215842
-= SORT MODEL (WEIGHT/VARIANCE) =-
before: after:
GMM 1: 0.000196 GMM 1: 0.002158
GMM 2: 0.000196 GMM 2: 0.000196
GMM 3: 0.000196 GMM 3: 0.000196
GMM 4: 0.000196 GMM 4: 0.000196
GMM 5: 0.002158 GMM 5: 0.000196
-=CHOOSE FIRST BACKGROUND/B MODEL (Threshold/T=0.700000) =-
B=1
total weight: 0.215842 > 0.700000 ? = N
B=2
total weight: 0.411881 > 0.700000 ? = N
B=3
total weight: 0.607921 > 0.700000 ? = N
B=4
total weight: 0.803960 > 0.700000 ? = N
=======================
***** iteration 1 *****
R:90 G:150 B:40
AND SO ON

Antworten (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by