Capsule shape detection in image - matlab

I have a binary image and I'd like to detect all approximately capsule-shaped regions like this (marked in red).
The size and orientation of desired regions is different for each image and unknown.
The desired ratio of height and width is between 3 and 5.
I don't know how to detect shape with unknown orientation and size, so I tried to use the ratio of the region's area and its bounding box's diagonal squared as criterion, but the result is not very good. Here's the code I use.
regions = bwconncomp(img);
sizes = cellfun(#numel, regions.PixelIdxList);
N = regions.NumObjects;
is_capsule = false(N, 1);
for k = 1:N
idx = regions.PixelIdxList{k};
[row, col] = ind2sub(sz, idx);
h = max(row) - min(row) + 1;
w = max(col) - min(col) + 1;
l = numel(idx) / (h^2 + w^2);
if l >= 0.1 && l <= 0.2
is_capsule(k) = 1;
end
end

Related

How to apply matched filter over surface of image instead of lines on 2d matrix: MATLAB

I have a 2d matrix and want to implement a matched filter on this matrix(image) of size 360x360
a=360x360
I tried to read about the matched filter but still unsure about it?
MY ATTEMPT(But this is not desired as this makes matched filter over lines and not on my image)
img=a;
s = 1.5; %sigma
L = 7;
theta = 0:15:165;
out = zeros(size(img));
m = max(ceil(3*s),(L-1)/2);
[x,y] = meshgrid(-m:m,-m:m);
for t = theta
t = t / 180 * pi;
u = cos(t)*x - sin(t)*y;
v = sin(t)*x + cos(t)*y;
N = (abs(u) <= 3*s) & (abs(v) <= L/2);
k = exp(-u.^2/(2*s.^2));
k = k - mean(k(N));
k(~N) = 0;
res = conv2(img,k,'same');
out = max(out,res);
end
imagesc(out)
My expectation was to get a more clear and bright image but I didn't get that
My expectation was to get a more clear and bright image of this coin but I didn't get that
I want to reduce noise and get a better image
I want to reduce noise and apply over image surface, not on lines get a better image

Simulating random walkers which can not collide into each other in Matlab

I have written a code to simulate the motion of circular particles in a 2d box. Whenever they move out of the box, I put them inside the box and near the wall. I want to add the diameter (2R) of particles in the code, which means when the distance between the center of two circles become less than 2R, they separate along the line connecting their centers so that the distance between the centers of the circles becomes equal to 2R.
Could anyone suggest a code to perevent the overlapping of particles?
This is my code in which overlap is not considered:
clear all
close all
l = 224; nn = 800; %number of particles
time = 1000; dd = 1;
x= l*rand(1,nn);
y= l*rand(1,nn);
for t = 1:time;
x= x + rand(1,nn)-0.5* ones(1,nn);
y=y+rand(1,nn)-0.5* ones (1,nn);
index = (x < 0); x(index) = abs(normrnd(0,1,1,nnz(index)));
index = (y < 0); y(index) = abs(normrnd(0,1,1,nnz(index)));
index = (x > l); x(index) = l-abs(normrnd(0,1,1,nnz(index)));
index = (y > l); y(index) = l-abs(normrnd(0,1,1,nnz(index)));
end
Here is some commented code which does what you want. Notably:
psize is some defined particle size for interaction.
point-to-point distances found using pdist2.
points that are too close are moved away from each other by some amount (dp times their current distances, if dp=1/2 then their x and y distances double) until there are no clashes.
See comments for details.
clear; close all;
l = 224; nn = 800; % number of particles
time = 100;
x = l*rand(1,nn); y = l*rand(1,nn);
psize = 2; % Particle size for interaction
dp = 0.1;
figure; hold on; axis([0 l 0 l]);
for t = 1:time;
% Random movement
movement = 2*rand(2,nn)-1;
x = x + movement(1,:);
y = y + movement(2,:);
index = (x < 0); x(index) = abs(normrnd(0,1,1,nnz(index)));
index = (y < 0); y(index) = abs(normrnd(0,1,1,nnz(index)));
index = (x > l); x(index) = l-abs(normrnd(0,1,1,nnz(index)));
index = (y > l); y(index) = l-abs(normrnd(0,1,1,nnz(index)));
% Particle interaction. Loop until there are no clashes. For
% robustness, some max iteration counter should be added!
numclash = 1;
while numclash > 0
dists = pdist2([x;y]', [x;y]'); % Distances between all particles
dists(dists < psize) = NaN; % Those too close are assigned NaN
tooclose = isnan(tril(dists,-1)); % All NaNs identified by logical
[clash1,clash2] = find(tooclose); % Get particles which are clashing
numclash = numel(clash1); % Get number of clashes
% All points where there was a clash, move away from each other
x(clash1) = x(clash1) + (x(clash1)-x(clash2))*dp;
x(clash2) = x(clash2) - (x(clash1)-x(clash2))*dp;
y(clash1) = y(clash1) + (y(clash1)-y(clash2))*dp;
y(clash2) = y(clash2) - (y(clash1)-y(clash2))*dp;
end
% Plot to visualise results. Colour fade from dark to bright green over time
scatter(x,y,'.','markeredgecolor',[0.1,t/time,0.4]);
drawnow;
end
hold off
Result:
Edit:
For a clearer diagram, you could initialise some colour matrix C = rand(nn,3); and plot using
scatter(x,y,[],C*(t/time),'.'); % the (t/time) factor makes it fade from dark to light
This would give each particle a different colour, which also fade from dark to light, rather than just fading from dark to light as before. The result would be something like this:

Shortest line between boundary points that passes through the centroid of a shape

lesion image
I have an irregularly shaped object in which I have to find the greatest and smallest diameter.
To find the greatest diameter, I extracted the boundary points and found the distances between all the points. I took the maximum distance amongst those distances which gave me my greatest diameter.
boundaries = bwboundaries(binaryImage);
numberOfBoundaries = size(boundaries, 1);
for blobIndex = 1 : numberOfBoundaries
thisBoundary = boundaries{blobIndex};
x = thisBoundary(:, 2); % x = columns.
y = thisBoundary(:, 1); % y = rows.
% Find which two boundary points are farthest from each other.
maxDistance = -inf;
for k = 1 : length(x)
distances = sqrt( (x(k) - x) .^ 2 + (y(k) - y) .^ 2 );
[thisMaxDistance, indexOfMaxDistance] = max(distances);
if thisMaxDistance > maxDistance
maxDistance = thisMaxDistance;
index1 = k;
index2 = indexOfMaxDistance;
end
end
I have attached my image containing the longest diameter.
I also need a line segment that passes through the centroid connecting the two boundary points whose length is shortest. When I try to find the shortest diameter by modifying the above code, to find min(distances), I am getting an error that says:
Error using griddedInterpolant
The coordinates of the input points must be finite values; Inf and NaN are not permitted.
What do I need to do to find the shortest "diameter" (that is, passing through the centroid) of this object?
it's possible to use a polar image like this:
lesion = imread('lesion.jpg');
bw = lesion > 100;
c = regionprops(bw,'Centroid');
c = c.Centroid;
% polar args
t = linspace(0,2*pi,361);
t(end) = [];
r = 0:ceil(sqrt(numel(bw)/4));
[tg,rg] = meshgrid(t,r);
[xg,yg] = pol2cart(tg,rg);
xoff = xg + c(1);
yoff = yg + c(2);
% polar image
pbw = interp2(double(bw),xoff,yoff,'nearest') == 1;
[~,radlen] = min(pbw,[],1);
radlen(radlen == 1) = max(r);
n = numel(radlen);
% add two edges of line to form diameter
diamlen = radlen(1:n/2) + radlen(n/2+1:n);
% find min diameter
[mindiam,tminidx1] = min(diamlen);
tmin = t(tminidx1);
rmin = radlen(tminidx1);
tminidx2 = tminidx1 + n/2;
xx = [xoff(radlen(tminidx1),tminidx1) xoff(radlen(tminidx2),tminidx2)];
yy = [yoff(radlen(tminidx1),tminidx1) yoff(radlen(tminidx2),tminidx2)];
% viz
figure;
subplot(121);
imshow(pbw);
title('polar image');
subplot(122);
imshow(bw);
hold on
plot(c(1),c(2),'or')
plot(xx,yy,'g')
legend('centroid','shortest diameter');
and the output is:

How to square the corners of a "rectangle" in a bw image with matlab

I have images of rectangles or deformed rectangles with rounded corners, like this:
or this:
is there a way to make the corners squared with matlab?
And then how can i get the coordinates of those new corners?
Thank you
Explanation
This problem is similar to the following question. My answer will be somehow similar to my answer there, with the relevant modifications.
we want to find the parallelogram corners which fits the most to the given shape.
The solution can be found by optimization, as follows:
find an initial guess for the 4 corners of the shape. This can be done by finding the boundary points with the highest curvature, and use kmean clustering to cluster them into 4 groups.
create a parallelogram given these 4 corners, by drawing a line between each pair of corresponding corners.
find the corners which optimize the Jaccard coefficient of the boundary image and the generated parallelogram map.
The optimization will done locally on each corner, in order to spare time.
Results
Initial corner guess (corners are marked in blue)
final results:
Code
main script
%reads image and binarize it
I = rgb2gray(imread('eA4ci.jpg')) > 50;
%finds boundry of largerst connected component
boundries = bwboundaries(I,8);
numPixels = cellfun(#length,boundries);
[~,idx] = max(numPixels);
B = boundries{idx};
%finds best 4 corners
[ corners ] = optimizeCorners(B);
%generate line mask given these corners, fills the result
linesMask = drawLines(size(I),corners,corners([2:4,1],:));
rectMask = imfill(linesMask,'holes');
%remove biggest CC from image, adds linesMask instead
CC = bwconncomp(I,8);
numPixels = cellfun(#numel,CC.PixelIdxList);
[~,idx] = max(numPixels);
res = I;
res(CC.PixelIdxList{idx}) = 0;
res = res | rectMask;
optimize corners function:
function [ corners] = optimizeCorners(xy)
%finds the corners which fits the most for this set of points
Y = xy(:,1);
X = xy(:,2);
%initial corners guess
corners = getInitialCornersGuess(xy);
boundriesIm = zeros(max(Y)+20,max(X)+20);
boundriesIm(sub2ind(size(boundriesIm),xy(:,1),xy(:,2))) = 1;
%R represents the search radius
R = 7;
%continue optimizing as long as there is no change in the final result
unchangedIterations = 0;
while unchangedIterations<4
for ii=1:4
%optimize corner ii
currentCorner = corners(ii,:);
bestCorner = currentCorner;
bestRes = calcEnergy(boundriesIm,corners);
cornersToEvaluate = corners;
for yy=currentCorner(1)-R:currentCorner(1)+R
for xx=currentCorner(2)-R:currentCorner(2)+R
cornersToEvaluate(ii,:) = [yy,xx];
res = calcEnergy(boundriesIm,cornersToEvaluate);
if res > bestRes
bestRes = res;
bestCorner = [yy,xx];
end
end
end
if isequal(bestCorner,currentCorner)
unchangedIterations = unchangedIterations + 1;
else
unchangedIterations = 0;
corners(ii,:) = bestCorner;
end
end
end
end
function res = calcEnergy(boundriesIm,corners)
%calculates the score of the corners list, given the boundries image.
%the result is acutally the jaccard index of the boundries map and the
%lines map
linesMask = drawLines(size(boundriesIm),corners,corners([2:4,1],:));
res = sum(sum(linesMask&boundriesIm)) / sum(sum(linesMask|boundriesIm));
end
get initial corners function:
function corners = getInitialCornersGuess(boundryPnts)
%calculates an initial guess for the 4 corners
%finds corners by performing kmeans on largest curvature pixels
[curvatureArr] = calcCurvature(boundryPnts, 5);
highCurv = boundryPnts(curvatureArr>0.3,:);
[~,C] = kmeans([highCurv(:,1),highCurv(:,2)],4);
%sorts the corners from top to bottom - preprocessing stage
C = int16(C);
corners = zeros(size(C));
%top left corners
topLeftInd = find(sum(C,2)==min(sum(C,2)));
corners(1,:) = C(topLeftInd,:);
%bottom right corners
bottomRightInd = find(sum(C,2)==max(sum(C,2)));
corners(3,:) = C(bottomRightInd,:);
%top right and bottom left corners
C([topLeftInd,bottomRightInd],:) = [];
topRightInd = find(C(:,2)==max(C(:,2)));
corners(4,:) = C(topRightInd,:);
bottomLeftInd = find(C(:,2)==min(C(:,2)));
corners(2,:) = C(bottomLeftInd,:);
end
function [curvatureArr] = calcCurvature(xy, halfWinSize)
%calculate the curvature of a list of points (xy) given a window size
%curvature calculation
curvatureArr = zeros(size(xy,1),1);
for t=1:halfWinSize
y = xy(t:halfWinSize:end,1);
x = xy(t:halfWinSize:end,2);
dx = gradient(x);
ddx = gradient(dx);
dy = gradient(y);
ddy = gradient(dy);
num = abs(dx .* ddy - ddx .* dy) + 0.000001;
denom = dx .* dx + dy .* dy + 0.000001;
denom = sqrt(denom);
denom = denom .* denom .* denom;
curvature = num ./ denom;
%normalizing
if(max(curvature) > 0)
curvature = curvature / max(curvature);
end
curvatureArr(t:halfWinSize:end) = curvature;
end
end
draw lines function:
function mask = drawLines(imgSize, P1, P2)
%generates a mask with lines, determine by P1 and P2 points
mask = zeros(imgSize);
P1 = double(P1);
P2 = double(P2);
for ii=1:size(P1,1)
x1 = P1(ii,2); y1 = P1(ii,1);
x2 = P2(ii,2); y2 = P2(ii,1);
% Distance (in pixels) between the two endpoints
nPoints = ceil(sqrt((x2 - x1).^2 + (y2 - y1).^2));
% Determine x and y locations along the line
xvalues = round(linspace(x1, x2, nPoints));
yvalues = round(linspace(y1, y2, nPoints));
% Replace the relevant values within the mask
mask(sub2ind(size(mask), yvalues, xvalues)) = 1;
end

Find Specified Color In Each Frame Of Real-time Video

I am newbie in Matlab.
My Boss gave me a task and i stuck in last step.
the task is:
user should specify a rectangle in snapshot of real-time video and then i should detect the mean color of that rectangle in each frame of video.
first user specify the boundaries of rectangle in a snapshot of video with this code:
test = getsnapshot(vid);
imwrite(test,'mytest.png','png');
testsnap = imread('mytest.png');
Rectpos=getrect;
then I calculate mean degree of each R , G , B :
bbx=boundingboxPixels(testsnap,Rectpos(1),Rectpos(2),Rectpos(3),Rectpos(4));
rRect=mean(bbx(:,:,1));
gRect=mean(bbx(:,:,2));
bRect=mean(bbx(:,:,3));
where boundingboxPixels method is like this :
function pixel_vals = boundingboxPixels(img, x_init, y_init, x_width, y_width)
if x_init > size(img,2)
error('x_init lies outside the bounds of the image.'); end
if y_init > size(img,1)
error('y_init lies outside the bounds of the image.'); end
if y_init+y_width > size(img,1) || x_init+x_width > size(img,2) || ...
x_init < 1 || y_init < 1
warning([...
'Given rectangle partially falls outside image. ',...
'Resizing rectangle...']);
end
x_min = max(1, uint16(x_init));
y_min = max(1, uint16(y_init));
x_max = min(size(img,2), x_min+uint16(x_width));
y_max = min(size(img,1), y_min+uint16(y_width));
x_range = x_min : x_max;
y_range = y_min : y_max;
Upper = img( x_range, y_min , :);
Left = img( x_min, y_range, :);
Right = img( x_max, y_range, :);
Lower = img( x_range, y_max , :);
pixel_vals = [...
Upper
permute(Left, [2 1 3])
permute(Right, [2 1 3])
Lower];
end
then get the calculated Mean of RGB color with a threshold from each frame of video:
tv=getdata(vid,1);//vid is real-time video
r=tv(:,:,1,1);
g=tv(:,:,2,1);
b=tv(:,:,3,1);
redVal = (r >= rRect-thr) & (r <= rRect+thr);
greenVal = (g >= gRect-thr) & (g <= gRect+thr);
blueVal = (b >= bRect-thr) & (b <= bRect+thr);
Now How Should I use the redVal , greenVal , blueVal to detect this color?
as Steffen said , problem solved by adding & between arrays