基于matlab CCF算法的图像质量评价

haidaowang

源代码

``
4 L  y9 @  {# A$ p0 a5 U. Y7 i* P- Wclear all
6 h3 g" m6 c0 Xclc
( Y! e8 \& `  O) Z+ i5 l5 I/ y
4 K, F* z5 @  r0 s# P7 [) _! Wim = imread(‘1.bmp’);

quality = CCF(im)
& m+ q% o1 a1 M1 \, f/ _; C9 m4 F3 @  Pfunction [quality] = CCF(im)
- I1 |) O% i( W9 C3 [
%-------------CCF_coloRFulness -----------------
imColor = double(im);


8 Z- ]5 B3 t/ j/ S: P4 E& g, W2 d

• R = imColor(:,:,1);
• G = imColor(:,:,2);
• B = imColor(:,:,3);
• RR = log(R+0.00001) - mean2(log(R+0.00001));
• GG = log(G+0.00001) - mean2(log(G+0.00001));
• BB = log(B+0.00001) - mean2(log(B+0.00001));
• alpha = RR-GG;
• beta = 0.5*(RR+GG)-BB;
• mu_alpha=mean(mean(alpha));
• mu_beta=mean(mean(beta));
• var_alpha=var(var(alpha));
• var_beta=var(var(beta));
• CCF_colorfulness=1000*((sqrt(var_alpha+var_beta)+0.3*sqrt(mu_alpha*mu_alpha+mu_beta*mu_beta))/85.59);
  

%-------------CCF_contrast----------------------
2 h' Z8 t9 d7 z* b6 ^9 F2 T7 ^

• im1=rgb2gray(im);
• CCF_contrast = CCFcontrast(im1);
  / Q/ {. P% W( _- x, W

! m! J6 Y( C( Q" F* B2 d' }& g
& n" J" w, s1 H%-------------CCF_FADE--------------------------

• CCF_FADE = FADE(im);
  6 H* X, W4 M. `3 H- S- R# }

; f& I; p$ S2 R* H/ x7 k%------------ normalization ------------------
  ]+ E" U) K/ X
7 ?2 B; b: h5 D0 [* m  p

• CCF_colorfulness=mapminmax(CCF_colorfulness,1,10);
• CCF_contrast=mapminmax(CCF_contrast,1,10);
• CCF_FADE=10-mapminmax(CCF_FADE,1,10);
  

8 n8 Z% G( M9 q/ e" B) F
8 [5 ?; J/ `8 p9 \! i) m( a! q* z. y7 U% ------------calculate image quality with coefficients---------------------

• c=[0.17593 0.61759 0.33988 ];
• quality = c(1)*(CCF_colorfulness) + c(2)*(CCF_contrast) + c(3)*(CCF_FADE)
  - g' }+ F6 R, b& Z  p

2 Z  t9 t: Y2 f' ]  W
" F5 U- a3 o: r- k* P# Z# lend


4 U( [4 S9 z1 ~# V# G

NNNei256

基于matlab CCF算法的图像质量评价

twel2e

可以借鉴一下
% Q& R% n  W- ?/ J0 U; n0 f
2 M( h' ^/ o! d2 B+ K2 J) k