Feature extraction of breast lesions

Question

Warid Islam 2020-8-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/582035-feature-extraction-of-breast-lesions

评论： Walter Roberson 2020-8-21

Hello,

I am trying to extract features from a segmented breast lesion. However I am getting an error when I try to use the regionprops function. Please find the code below:

clc;clear;close all
% Specify the folder where the files live.
myFolder = 'C:\Users\morteza\Desktop\Udemy Courses\Deep Learning A-Z\P16-Convolutional-Neural-Networks (1)\Convolutional_Neural_Networks\Images_Malignant\training_set\Benign\Images';
% Check to make sure that folder actually exists.  Warn user if it doesn't.
if ~isfolder(myFolder)
    errorMessage = sprintf('Error: The following folder does not exist:\n%s\nPlease specify a new folder.', myFolder);
    uiwait(warndlg(errorMessage));
    myFolder = uigetdir(); % Ask for a new one.
    if myFolder == 0
         % User clicked Cancel
         return;
    end
end
% Get a list of all files in the folder with the desired file name pattern.
filePattern = fullfile(myFolder, '*.jpg'); % Change to whatever pattern you need.
theFiles = dir(filePattern);
for k = 1 : length(theFiles)
    baseFileName = theFiles(k).name;
    fullFileName = fullfile(theFiles(k).folder, baseFileName); 
    fprintf(1, 'Now reading %s\n', fullFileName); 
    % Now do whatever you want with this file name,
    % such as reading it in as an image array with imread()
      r = im2double((imread(fullFileName)));
try
    r=rgb2gray(r);
end
J = imnoise(r,'salt & pepper',0.02);
c=medfilt2(J);
b= adapthisteq(c);
% b = medfilt2(r1);
hy = fspecial('sobel');
hx = hy';
Iy = imfilter(double(b), hy, 'replicate');
Ix = imfilter(double(b), hx, 'replicate');
gradmag = sqrt(Ix.^2 + Iy.^2);
L = watershed(gradmag);
Lrgb = label2rgb(L);
se = strel('disk', 20);
Io = imopen(b, se);
Ie = imerode(b, se);
Iobr = imreconstruct(Ie, b);
Ioc = imclose(Io, se);
Iobrd = imdilate(Iobr, se);
Iobrcbr = imreconstruct(imcomplement(Iobrd), imcomplement(Iobr));
Iobrcbr = imcomplement(Iobrcbr);
fgm = imregionalmax(Iobrcbr);
% figure
Y = imclearborder(fgm);
stats1 = regionprops(Y,'Area','Circularity','Solidity','Eccentricity', 'EquivDiameter','EulerNumber','MajorAxisLength','MinorAxisLength','Perimeter');
area=stats1.Area;
circularity=stats1.Circularity;
solidity=stats1.Solidity;
eccentricity=stats1.Eccentricity;
equivdiameter=stats1.EquivDiameter;
eulernumber=stats1.EulerNumber;
major=stats1.MajorAxisLength;
minor=stats1.MinorAxisLength;
perimeter=stats1.Perimeter;
p(k,:)=[area,circularity,solidity,eccentricity,equivdiameter,eulernumber,major,minor,perimeter];

I get the following error message:

Insufficient number of outputs from right hand side of equal sign to satisfy assignment.

Error in gf (line 84)

area=stats1.Area;

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2020-8-20

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/582035-feature-extraction-of-breast-lesions#answer_482570

在 MATLAB Online 中打开

area = vertcat(stats1.Area);
circularity = vertcat(stats1.Circularity);
solidity = vertcat(stats1.Solidity);
eccentricity = vertcat(stats1.Eccentricity);
equivdiameter = vertcat(stats1.EquivDiameter);
eulernumber = vertcat(stats1.EulerNumber);
major = vertcat(stats1.MajorAxisLength);
minor = vertcat(stats1.MinorAxisLength);
perimeter = vertcat(stats1.Perimeter);

However, your line

p(k,:)=[area,circularity,solidity,eccentricity,equivdiameter,eulernumber,major,minor,perimeter];

is going to have problems. You are going to need to turn p into a cell array.

The difficulty you are having is that you are assuming that your Y only has a single region in it. Now, it might be the case that you want to use bwareafilt() to pick out only the largest of the regions, but unless you do something like that, you are going to be getting multiple regions.

6 个评论
显示 4更早的评论隐藏 4更早的评论

Warid Islam 2020-8-21

Hi Walter,

The problem is if i try to use other feature extraction method such as the Haralick Texture Feature Extractiom, Linear Binary pattern, etc. then the algorithm works. There seems to be a problem while using regionprops. Do you think I am missing something in the code?

Walter Roberson 2020-8-21

No.

regionprops will find any region that exists, even just single pixel. If it returns empty information then the data you passed to it was empty or you passed Connected Component that was empty or you passed Label information to it that was empty.

请先登录，再进行评论。