Welcome to this blog
Advantages of bloggers:Blog content should be as thoughtful and logical as possible for the convenience of readers.
Motto:He who travels a hundred miles is half as good as ninety.
The directory of this article is as follows:
Table of Contents
1 Overview
2 Operation results
?
3 References
4 Matlab code implementation
1 Overview
[Image Segmentation] Image detection (segmentation, feature extraction), measurement and filtering of various features (area, etc.)
This article provides a beginner-friendly tutorial designed to demonstrate image detection (segmentation, feature extraction) as well as measurement and filtering of various features (such as area) (extracting only certain objects).
First, the tutorial shows how to find all objects (coins) in an image, then filter the results based on a specified diameter to filter out objects of a specific diameter. A simple example demonstrates the basic concepts of thresholding, marking, and region attributes.
This tutorial is a great starting point for students who are new to MATLAB’s image processing capabilities. They can use this tutorial to deepen their understanding of basic concepts and techniques before diving into more complex algorithms.
In order to complete this tutorial, you need to install the Image Processing Toolbox, as it demonstrates some of the functionality provided by the toolbox. At the same time, the tutorial uses a sample image called “coin” that comes with the toolbox as a demonstration object.
The beauty of this tutorial is that it provides an intuitive and practical way to help beginners understand how to use MATLAB to process and analyze images. By learning how to perform image segmentation, feature extraction, and filtering, readers will benefit from these basic concepts and be able to apply them to solve more complex image processing problems. This tutorial is a great starting point for students, researchers, and engineers interested in further exploring the field of image processing.
2 Running results
Part of the code:
% Read in a standard MATLAB demo image of coins (US nickles and dimes, which are 5 cent and 10 cent coins). This image ships with MATLAB.
baseFileName = 'coins.png';
folder = fileparts(which(baseFileName)); % Determine where demo folder is (works with all versions).
fullFileName = fullfile(folder, baseFileName);
fprintf('Full File Name = "%s".\
', fullFileName);
if ~exist(fullFileName, 'file')
% It doesn't exist in the current folder.
% Look on the search path.
if ~exist(baseFileName, 'file')
% It doesn't exist on the search path either.
% Alert user that we can't find the image.
warningMessage = sprintf('Error: the input image file\
%s\
was not found.\
Click OK to exit the demo.', fullFileName);
uiwait(warndlg(warningMessage));
fprintf(1, 'Finished running BlobsDemo.m.\
');
return;
end
% Found it on the search path. Construct the file name.
fullFileName = baseFileName; % Note: don't prepend the folder.
end
% If we get here, we should have found the image file.
originalImage = imread(fullFileName);
% Check to make sure that it is grayscale, just in case the user substituted their own image.
[rows, columns, numberOfColorChannels] = size(originalImage);
if numberOfColorChannels > 1
promptMessage = sprintf('Your image file has %d color channels.\
This demo was designed for grayscale images.\
Do you want me to convert it to grayscale for you so you can continue?', numberOfColorChannels);
button = questdlg(promptMessage, 'Continue', 'Convert and Continue', 'Cancel', 'Convert and Continue');
if strcmp(button, 'Cancel')
fprintf(1, 'Finished running BlobsDemo.m.\
');
return;
end
% Do the conversion using standard book formula
originalImage = rgb2gray(originalImage);
end
% Display the grayscale image.
subplot(3, 3, 1);
imshow(originalImage);
% Maximize the figure window.
hFig1 = gcf;
hFig1.Units = 'normalized';
hFig1.WindowState = 'maximized'; % Go to full screen.
hFig1.NumberTitle = 'off'; % Get rid of "Figure 1"
hFig1.Name = 'Demo by Image Analyst'; % Put this into title bar.
% Force it to display RIGHT NOW (otherwise it might not display until it's all done, unless you've stopped at a breakpoint.)
drawnow;
caption = sprintf('Original "coins" image showing\
6 nickels (the larger coins) and 4 dimes (the smaller coins).');
title(caption, 'FontSize', captionFontSize);
axis('on', 'image'); % Make sure image is not artificially stretched because of screen's aspect ratio.
% Just for fun, let's get its histogram and display it.
[pixelCount, grayLevels] = imhist(originalImage);
subplot(3, 3, 2);
bar(pixelCount);
title('Histogram of original image', 'FontSize', captionFontSize);
xlim([0 grayLevels(end)]); % Scale x axis manually.
grid on;
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Threshold the image to get a binary image (only 0's and 1's) of class "logical."
% Method #1: using im2bw()
% normalizedThresholdValue = 0.4; % In range 0 to 1.
% thresholdValue = normalizedThresholdValue * max(max(originalImage)); % Gray Levels.
% binaryImage = im2bw(originalImage, normalizedThresholdValue); % One way to threshold to binary
% Method #2: using a logical operation.
thresholdValue = 100;
binaryImage = originalImage > thresholdValue; % Bright objects will be chosen if you use >.
% ========== IMPORTANT OPTION ===================================== =======================
% Use < if you want to find dark objects instead of bright objects.
% binaryImage = originalImage <thresholdValue; % Dark objects will be chosen if you use <.
% Do a "hole fill" to get rid of any background pixels or "holes" inside the blobs.
binaryImage = imfill(binaryImage, 'holes');
% Show the threshold as a vertical red bar on the histogram.
hold on;
maxYValue = ylim;
line([thresholdValue, thresholdValue], maxYValue, 'Color', 'r');
% Place a text label on the bar chart showing the threshold.
annotationText = sprintf('Thresholded at %d gray levels', thresholdValue);
% For text(), the x and y need to be of the data class "double" so let's cast both to double.
text(double(thresholdValue + 5), double(0.5 * maxYValue(2)), annotationText, 'FontSize', 10, 'Color', [0 .5 0]);
text(double(thresholdValue - 70), double(0.94 * maxYValue(2)), 'Background', 'FontSize', 10, 'Color', [0 0 .5]);
text(double(thresholdValue + 50), double(0.94 * maxYValue(2)), 'Foreground', 'FontSize', 10, 'Color', [0 0 .5]);
% Display the binary image.
subplot(3, 3, 3);
imshow(binaryImage);
title('Binary Image, obtained by thresholding', 'FontSize', captionFontSize);
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Identify individual blobs by seeing which pixels are connected to each other. This is called "Connected Components Labeling".
% Each group of connected pixels will be given a label, a number, to identify it and distinguish it from the other blobs.
% Do connected components labeling with either bwlabel() or bwconncomp().
[labeledImage, numberOfBlobs] = bwlabel(binaryImage, 8); % Label each blob so we can make measurements of it
% labeledImage is an integer-valued image where all pixels in the blobs have values of 1, or 2, or 3, or ... etc.
subplot(3, 3, 4);
imshow(labeledImage, []); % Show the gray scale image.
title('Labeled Image, from bwlabel()', 'FontSize', captionFontSize);
drawnow;
% Let's assign each blob a different color to visually show the user the distinct blobs.
coloredLabels = label2rgb (labeledImage, 'hsv', 'k', 'shuffle'); % pseudo random color labels
% coloredLabels is an RGB image. We could have applied a colormap instead (but only with R2014b and later)
subplot(3, 3, 5);
imshow(coloredLabels);
axis image; % Make sure image is not artificially stretched because of screen's aspect ratio.
caption = sprintf('Pseudo colored labels, from label2rgb().\
Blobs are numbered from top to bottom, then from left to right.');
title(caption, 'FontSize', captionFontSize);
%================================================== ================================================== ================================================== =
% MAIN PART IS RIGHT HERE!!!
% Get all the blob properties.
props = regionprops(labeledImage, originalImage, 'all');
% Or, if you want, you can ask for only a few specific measurements. This will be faster since we don't have to compute everything.
% props = regionprops(labeledImage, originalImage, 'MeanIntensity', 'Area', 'Perimeter', 'Centroid', 'EquivDiameter');
numberOfBlobs = numel(props); % Will be the same as we got earlier from bwlabel().
%================================================== ================================================== ================================================== =
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% PLOT BOUNDARIES.
% Plot the borders of all the coins on the original grayscale image using the coordinates returned by bwboundaries().
% bwboundaries() returns a cell array, where each cell contains the row/column coordinates for an object in the image.
subplot(3, 3, 6);
imshow(originalImage);
title('Outlines, from bwboundaries()', 'FontSize', captionFontSize);
axis('on', 'image'); % Make sure image is not artificially stretched because of screen's aspect ratio.
% Here is where we actually get the boundaries for each blob.
boundaries = bwboundaries(binaryImage); % Note: this is a cell array with several boundaries -- one boundary per cell.
% boundaries is a cell array - one cell for each blob.
% In each cell is an N-by-2 list of coordinates in a (row, column) format. Note: NOT (x,y).
% Column 1 is rows, or y. Column 2 is columns, or x.
numberOfBoundaries = size(boundaries, 1); % Count the boundaries so we can use it in our for loop
% Here is where we actually plot the boundaries of each blob in the overlay.
hold on; % Don't let boundaries blow away the displayed image.
for k = 1 : numberOfBoundaries
thisBoundary = boundaries{k}; % Get boundary for this specific blob.
x = thisBoundary(:,2); % Column 2 is the columns, which is x.
y = thisBoundary(:,1); % Column 1 is the rows, which is x.
plot(x, y, 'r-', 'LineWidth', 2); % Plot boundary in red.
end
hold off;
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Print out the measurements to the command window, and display blob numbers on the image.
textFontSize = 14; % Used to control size of "blob number" labels put atop the image.
% Print header line in the command window.
fprintf(1,'Blob # Mean Intensity Area Perimeter Centroid Diameter\
');
% Extract all the mean diameters into an array.
% The "diameter" is the "Equivalent Circular Diameter", which is the diameter of a circle with the same number of pixels as the blob.
% Enclosing in brackets is a nice trick to concatenate all the values from all the structure fields (every structure in the props structure array).
blobECD = [props.EquivDiameter];
% Loop over all blobs printing their measurements to the command window.
for k = 1 : numberOfBlobs % Loop through all blobs.
% Find the individual measurements of each blob. They are field of each structure in the props strucutre array.
% You could use the bracket trick (like with blobECD above) OR you can get the value from the field of this particular structure.
% I'm showing you both ways and you can use the way you like best.
meanGL = props(k).MeanIntensity; % Get average intensity.
blobArea = props(k).Area; % Get area.
blobPerimeter = props(k).Perimeter; % Get perimeter.
blobCentroid = props(k).Centroid; % Get centroid one at a time
% Now do the printing of this blob's measurements to the command window.
fprintf(1,'#- .1f .1f %8.1f %8.1f %8.1f % 8.1f\
', k, meanGL, blobArea, blobPerimeter, blobCentroid, blobECD(k));
% Put the "blob number" labels on the grayscale image that is showing the red boundaries on it.
text(blobCentroid(1), blobCentroid(2), num2str(k), 'FontSize', textFontSize, 'FontWeight', 'Bold', 'HorizontalAlignment', 'center', 'VerticalAlignment', 'middle');
end
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Now, I'll show you a way to get centroids into an N -by-2 array directly from props,
% rather than accessing them as a field of the props strcuture array.
% We can get the centroids of ALL the blobs into 2 arrays,
% one for the centroid x values and one for the centroid y values.
allBlobCentroids = vertcat(props.Centroid); % A 10 row by 2 column array of (x,y) centroid coordinates.
centroidsX = allBlobCentroids(:, 1); % Extract out the centroid x values into their own vector.
centroidsY = allBlobCentroids(:, 2); % Extract out the centroid y values into their own vector.
% Put the labels on the rgb labeled image also.
subplot(3, 3, 5);
for k = 1 : numberOfBlobs % Loop through all blobs.
% Place the blob label number at the centroid of the blob.
text(centroidsX(k), centroidsY(k), num2str(k), 'FontSize', textFontSize, 'FontWeight', 'Bold', 'HorizontalAlignment', 'center', 'VerticalAlignment', 'middle');
end
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Now I'll demonstrate how to select certain blobs based using the ismember() function and extract them into new subimages.
% Let's say that we wanted to find only those blobs
% with an intensity between 150 and 220 and an area less than 2000 pixels.
% This would give us the three brightest dimes (the smaller coin type).
allBlobIntensities = [props.MeanIntensity];
allBlobAreas = [props.Area];
subplot(3, 3, 7);
histogram(allBlobAreas);
% Get a list of the blobs that meet our criteria and we need to keep.
% These will be logical indices - lists of true or false depending on whether the feature meets the criteria or not.
% for example [1, 0, 0, 1, 1, 0, 1, .....]. Elements 1, 4, 5, 7, ... are true, others are false.
allowableIntensityIndexes = (allBlobIntensities > 150) & amp; (allBlobIntensities < 220);
allowableAreaIndexes = allBlobAreas < 2000; % Take the small objects.
% Now let's get actual indexes, rather than logical indexes, of the features that meet the criteria.
% for example [1, 4, 5, 7, .....] to continue using the example from above.
keeperIndexes = find(allowableIntensityIndexes & amp; allowableAreaIndexes);
% Extract only those blobs that meet our criteria, and
% eliminate those blobs that don't meet our criteria.
% Note how we use ismember() to do this. Result will be an image - the same as labeledImage but with only the blobs listed in keeperIndexes in it.
keeperBlobsImage = ismember(labeledImage, keeperIndexes);
% Re-label with only the keeper blobs kept.
labeledDimeImage = bwlabel(keeperBlobsImage, 8); % Label each blob so we can make measurements of it
% Now we're done. We have a labeled image of blobs that meet our specified criteria.
subplot(3, 3, 7);
imshow(labeledDimeImage, []);
axis image;
title('"Keeper" blobs (3 brightest dimes in a re-labeled image)', 'FontSize', captionFontSize);
elapsedTime = toc;
fprintf('Blob detection and measurement took %.3f seconds.\
', elapsedTime)
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Plot the centroids in the overlay above the original image in the upper left axes.
% Dimes will have a red cross, nickels will have a blue X.
message = sprintf('Now I will plot the centroids over the original image in the upper left.\
Please look at the upper left image.');
reply = questdlg(message, 'Plot Centroids?', 'OK', 'Cancel', 'Cancel');
% Note: reply will = '' for Upper right X, 'OK' for OK, and 'Cancel' for Cancel.
if strcmpi(reply, 'Cancel')
return;
end
subplot(3, 3, 1);
hold on; % Don't blow away image.
for k = 1 : numberOfBlobs % Loop through all keeper blobs.
% Identify if blob #k is a dime or nickel.
itsADime = allBlobAreas(k) < 2200; % Dimes are small.
if itsADime
% Plot dimes with a red + .
plot(centroidsX(k), centroidsY(k), 'r + ', 'MarkerSize', 15, 'LineWidth', 2);
else
% Plot nickels with a blue x.
plot(centroidsX(k), centroidsY(k), 'bx', 'MarkerSize', 15, 'LineWidth', 2);
end
end
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Now use the keeper blobs as a mask on the original image so we will get a masked gray level image.
% This will keep the regions in the mask as original but erase (blacken) everything else (outside of the mask regions).
% This will let us display the original image in the regions of the keeper blobs.
maskedImageDime = originalImage; % Simply a copy at first.
maskedImageDime(~keeperBlobsImage) = 0; % Set all non-keeper pixels to zero.
subplot(3, 3, 8);
imshow(maskedImageDime);
axis image;
title('Only the 3 brightest dimes from the original image', 'FontSize', captionFontSize);
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Now let's get the nickels (the larger coin type).
keeperIndexes = find(allBlobAreas > 2000); % Take the larger objects.
% Note how we use ismember to select the blobs that meet our criteria. Get a binary image with only nickel regions present.
nickelBinaryImage = ismember(labeledImage, keeperIndexes);
% Let's get the nickels from the original grayscale image, with the other non-nickel pixels blackened.
% In other words, we will create a "masked" image.
maskedImageNickel = originalImage; % Simply a copy at first.
maskedImageNickel(~nickelBinaryImage) = 0; % Set all non-nickel pixels to zero.
subplot(3, 3, 9);
imshow(maskedImageNickel, []);
axis image;
title('Only the nickels from the original image', 'FontSize', captionFontSize);
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% WE'RE BASICALLY DONE WITH THE DEMO NOW.
elapsedTime = toc;
% Alert user that the demo is done and give them the option to save an image.
message = sprintf('Done making measurements of the features.\
\
Elapsed time = %.2f seconds.', elapsedTime);
message = sprintf('%s\
\
Check out the figure window for the images.\
Check out the command window for the numerical results.', message);
message = sprintf('%s\
\
Do you want to save the pseudo-colored image?', message);
reply = questdlg(message, 'Save image?', 'Yes', 'No', 'No');
% Note: reply will = '' for Upper right X, 'Yes' for Yes, and 'No' for No.
if strcmpi(reply, 'Yes')
% Ask user for a filename.
FilterSpec = {'*.PNG', 'PNG Images (*.png)'; '*.tif', 'TIFF images (*.tif)'; '*.*', 'All Files (*.*)' };
DialogTitle = 'Save image file name';
% Get the default filename. Make sure it's in the folder where this m-file lives.
% (If they run this file but the cd is another folder then pwd will show that folder, not this one.
thisFile = mfilename('fullpath');
[thisFolder, baseFileName, ext] = fileparts(thisFile);
DefaultName = sprintf('%s/%s.tif', thisFolder, baseFileName);
[fileName, specifiedFolder] = uiputfile(FilterSpec, DialogTitle, DefaultName);
if fileName ~= 0
% Parse what they actually specified.
[folder, baseFileName, ext] = fileparts(fileName);
% Create the full filename, making sure it has a tif filename.
fullImageFileName = fullfile(specifiedFolder, [baseFileName '.tif']);
% Save the labeled image as a tif image.
imwrite(uint8(coloredLabels), fullImageFileName);
% Just for fun, read image back into the imtool utility to demonstrate that tool.
tifimage = imread(fullImageFileName);
imtool(tifimage, []);
end
end
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% OPTIONAL : CROP EACH COIN OUT TO A SEPARATE SUB-IMAGE ON A NEW FIGURE.
message = sprintf('Would you like to crop out each coin to individual images?');
reply = questdlg(message, 'Extract Individual Images?', 'Yes', 'No', 'Yes');
% Note: reply will = '' for Upper right X, 'Yes' for Yes, and 'No' for No.
if strcmpi(reply, 'Yes')
% Maximize the figure window.
hFig2 = figure; % Create a new figure window.
hFig2.Units = 'normalized';
hFig2.WindowState = 'maximized'; % Go to full screen.
hFig2.NumberTitle = 'off'; % Get rid of "Figure 1"
hFig2.Name = 'Demo by Image Analyst'; % Put this into title bar.
for k = 1 : numberOfBlobs % Loop through all blobs.
% Find the bounding box of each blob.
thisBlobsBoundingBox = props(k).BoundingBox; % Get list of pixels in current blob.
% Extract out this coin into it's own image.
subImage = imcrop(originalImage, thisBlobsBoundingBox);
% Determine if it's a dime (small) or a nickel (large coin).
if props(k).Area > 2200
coinType = 'nickel';
else
coinType = 'dime';
end
% Display the image with informative caption.
subplot(3, 4, k);
imshow(subImage);
caption = sprintf('Coin #%d is a %s.\
Diameter = %.1f pixels\
Area = %d pixels', ...
k, coinType, blobECD(k), props(k).Area);
title(caption, 'FontSize', textFontSize);
end
%------------------------------------------------- -------------------------------------------------- -------------------------------------------------- -
% Display the MATLAB "peaks" logo.
logoSubplot = subplot(3, 4, 11:12);
caption = sprintf('A MATLAB Tutorial by ImageAnalyst');
text(0.5,1.15, caption, 'Color','r', 'FontSize', 18, 'FontWeight','b', 'HorizontalAlignment', 'center', 'VerticalAlignment', 'middle') ;
positionOfLowerRightPlot = get(logoSubplot, 'position');
L = 40*membrane(1,25);
logoax = axes('CameraPosition', [-193.4013, -265.1546, 220.4819],...
'Box', 'off', ...
'CameraTarget',[26, 26, 10], ...
'CameraUpVector',[0, 0, 1], ...
'CameraViewAngle',9.5, ...
'DataAspectRatio', [1, 1, .9],...
'Position', positionOfLowerRightPlot, ...
'Visible','off', ...
'XLim',[1, 51], ...
'YLim',[1, 51], ...
'ZLim',[-13, 40], ...
'parent', gcf);
axis(logoSubplot, 'off');
s = surface(L, ...
'EdgeColor','none', ...
'FaceColor',[0.9, 0.2, 0.2], ...
'FaceLighting','phong', ...
'AmbientStrength',0.3, ...
'DiffuseStrength',0.6, ...
'Clipping','off',...
'BackFaceLighting','lit', ...
'SpecularStrength',1, ...
'SpecularColorReflectance',1, ...
'SpecularExponent',7, ...
'Tag','TheMathWorksLogo', ...
'parent',logoax);
l1 = light('Position',[40, 100, 20], ...
% OPTIONAL : CROP EACH COIN OUT TO A SEPARATE SUB-IMAGE ON A NEW FIGURE.
message = sprintf(‘Would you like to crop out each coin to individual images?’);
reply = questdlg(message, ‘Extract Individual Images?’, ‘Yes’, ‘No’, ‘Yes’);
% Note: reply will = ” for Upper right X, ‘Yes’ for Yes, and ‘No’ for No.
if strcmpi(reply, ‘Yes’)
% Maximize the figure window.
hFig2 = figure; % Create a new figure window.
hFig2.Units = ‘normalized’;
hFig2.WindowState = ‘maximized’; % Go to full screen.
hFig2.NumberTitle = ‘off’; % Get rid of “Figure 1”
hFig2.Name = ‘Demo by Image Analyst’; % Put this into title bar.
for k = 1 : numberOfBlobs % Loop through all blobs.
% Find the bounding box of each blob.
thisBlobsBoundingBox = props(k).BoundingBox; % Get list of pixels in current blob.
% Extract out this coin into it’s own image.
subImage = imcrop(originalImage, thisBlobsBoundingBox);
% Determine if it’s a dime (small) or a nickel (large coin).
if props(k).Area > 2200
coinType = ‘nickel’;
else
coinType = ‘dime’;
end
% Display the image with informative caption.
subplot(3, 4, k);
imshow(subImage);
caption = sprintf(‘Coin #%d is a %s.\
Diameter = %.1f pixels\
Area = %d pixels’, …
k, coinType, blobECD(k), props(k).Area);
title(caption, ‘FontSize’, textFontSize);
end
%————————————————- ————————————————– ————————————————– –
% Display the MATLAB “peaks” logo.
logoSubplot = subplot(3, 4, 11:12);
caption = sprintf(‘A MATLAB Tutorial by ImageAnalyst’);
text(0.5,1.15, caption, ‘Color’,’r’, ‘FontSize’, 18, ‘FontWeight’,’b’, ‘HorizontalAlignment’, ‘center’, ‘VerticalAlignment’, ‘middle’) ;
positionOfLowerRightPlot = get(logoSubplot, ‘position’);
L = 40*membrane(1,25);
logoax = axes(‘CameraPosition’, [-193.4013, -265.1546, 220.4819],…
‘Box’, ‘off’, …
‘CameraTarget’,[26, 26, 10], …
‘CameraUpVector’,[0, 0, 1], …
‘CameraViewAngle’,9.5, …
‘DataAspectRatio’, [1, 1, .9],…
‘Position’, positionOfLowerRightPlot, …
3 References
Some of the content in the article is quoted from the Internet. The source will be indicated or cited as a reference. It is inevitable that there will be some unfinished information. If there is anything inappropriate, please feel free to contact us to delete it.
[1] Ma Yin. Image feature extraction and recognition based on CCD [D]. Northeastern University, 2012. DOI: 10.7666/d.J0120301.
[2] Wang Niu, Kang Huiying. Optimization algorithm for ship feature segmentation and extraction based on image detection [J]. Ship Science and Technology, 2018(4X):3.DOI:CNKI:SUN:JCKX.0.2018-08-049.
[3] Yin Cong. Color image face detection and feature extraction authentication [J]. Information Technology and Informatization, 2009. DOI: JournalArticle/5af35bd8c095d718d80b8d86.
[4] Luo Wenhui, Wang Sanwu. Two-step segmentation method of water meter images based on area and structural features [J]. Journal of Wuhan University of Technology: Information and Management Engineering Edition, 2006, 28(5):4.DOI:10.3963/j.issn .1007-144X.2006.05.014.