error-different values for same image

조회 수: 3 (최근 30일)
kash
kash 2012년 1월 21일
I am performing feature extraction,i have 80 images in data base,i am extracting 3 features,and have saved ,them,next am taking first image from that 80 images and have extracted same features .and have saved it.
wen i compare i get different same values for 2 feature extraction and for third feature extraction ,i get different values
the feature extraction aye
two colours and one texture
the code for texture is
g=rgb2gray(I);
rgb=mean(mean(I));
r1=rgb(:,:,1);
g1=rgb(:,:,2);
b1=rgb(:,:,3);
std1=std2(g);
entropy1=entropy(g);
tres=graythresh(g);
glcm=graycomatrix(g);
F=graycoprops(glcm,{'Contrast','Homogeneity','Correlation','Energy'});%graycoprops(glcm, properties)
%calculates the statistics specified in properties from the gray-level co-occurence matrix glcm
contrast=F.Contrast;
homogeneity=F.Homogeneity;
correlation=F.Correlation;
energy=F.Energy;
final_tex=[contrast,homogeneity,correlation,energy];
final_textq=final_tex;
save('final_text_Query', 'final_textq');
FinalOut2_Query = final_tex
please help
  댓글 수: 1
Walter Roberson
Walter Roberson 2012년 2월 4일
You somehow left out your question.

댓글을 달려면 로그인하십시오.

답변 (2개)

Walter Roberson
Walter Roberson 2012년 2월 4일
There is no reason to expect that different features will always give the same value. Feature extraction uses multiple features. When you do CBIR you match against the multiple features and come up with a majority or "most matches" or consensus decision knowing that not every feature will match.

Image Analyst
Image Analyst 2012년 2월 4일
Either your image is not the same like you thought it was, or you ran a different piece of code on the same image. Because running the same set of code on the same image will give the same result, unless you somehow used a function call to a random number generator (like rand, randi, randperm, etc.).

Community Treasure Hunt

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

Start Hunting!

Translated by