How can I fix this code to work?

Question

Mohammad Farhad Aryan 2020-4-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/514879-how-can-i-fix-this-code-to-work

回答： Asvin Kumar 2020-4-7

Capture.JPG

I am using the following code to extract mean feature and texture features using GLCM for every lesion in the input image and append them to an existing matrix then save it into a csv file. The code works when there is only one record in the matrix created but if there are more one record in the matrix created then it gives the following error.

'The size of the right hand side did not match the size of the indexing on the left hand side'.

clc;
clear all;
I = imread('Capture.JPG');
gray_image = rgb2gray(I);
binary_image = gray_image > 40;
imshow(binary_image);
roundBlobs = regionprops(binary_image, 'BoundingBox');
for k = 1 : length(roundBlobs)
    thisBB = roundBlobs(k).BoundingBox;
    crop_blob=imcrop(I,[thisBB(1),thisBB(2),thisBB(3),thisBB(4)]);
    resize_blob=imresize(crop_blob,[thisBB(3) thisBB(4)]);
    figure,imshow(resize_blob);
    
    % Calculate mean of color spaces
    redChannel = resize_blob(:, :, 1);
    greenChannel = resize_blob(:, :, 2);
    blueChannel = resize_blob(:, :, 3);
    % Get mean of R, G, and B
    meanR = mean2(redChannel);
    meanG = mean2(greenChannel);
    meanB = mean2(blueChannel); 
    
    HSVimage = rgb2hsv(resize_blob);
    H_component = HSVimage(:,:,1);
    S_component = HSVimage(:,:,2);
    I_component = HSVimage(:,:,3);
    % Get mean of Hue, Saturation, and Value
    meanH = mean(H_component(:));
    meanS = mean(S_component(:));
    meanV = mean(I_component(:));
    
    YCbCrimage = rgb2ycbcr(resize_blob);
    Y_component = YCbCrimage(:,:,1);
    Cb_component = YCbCrimage(:,:,2);
    Cr_component = YCbCrimage(:,:,3);
    % Get mean of Y, Cb, and Cr
    meanY = mean2(Y_component);
    meanCb = mean2(Cb_component);
    meanCr = mean2(Cr_component);
    Lab = rgb2lab(resize_blob);
    L = Lab(:,:,1);
    a = Lab(:,:,2);
    b = Lab(:,:,3);
    % Get mean of L, a, and b
    meanL = mean2(L);
    meanLA = mean2(a);
    meanLB = mean2(b);    
    
    % Calculate GLCM and extract texture features
    gray_blob = rgb2gray(resize_blob);
    offsets = [0 1; -1 1;-1 0;-1 -1];
    glcm = graycomatrix(gray_blob, 'GrayLimits', [], 'Offset',offsets, 'Symmetric', true);
    stats = graycoprops(glcm);
    avgCont = mean(stats.Contrast);
    avgCorrel = mean(stats.Correlation);
    avgEnergy = mean(stats.Energy);
    avgHomogen = mean(stats.Homogeneity);
    Feats = [avgCont, avgCorrel, avgEnergy, avgHomogen, meanR, meanG, meanB, meanH, meanS, meanV, meanY, meanCb, meanCr, meanL, meanLA, meanLB];
    glcm_data(k, :) = Feats(:);
end
m = matfile('gl_data.mat', 'Writable', true);
if isprop(m, 'gl_data')
    s = size(m, 'gl_data');
    m.gl_data(s(1)+1, :) = glcm_data;    
else
    m.gl_data = glcm_data;
end
load('gl_data.mat');
writematrix(gl_data, 'gl_data.csv');

I want to be able to append the extracted features for one lesion or multiple lesions to the existing matrix. For example if the input image has one lesion then the extracted data is only one record or if the input image has multiple lesions then the extracted data consists of multiple records so in both cases, I have to be able to append them to an existing matrix.

Any help is greatly appreciated!