I am trying to follow this MATLAB example.
Please see Step 2, the example used a 128 * 27 matrix M2, and use affine transform to scale and rotate M2, the scale factor is 2.5. However, I expect the size of the result should be 67.5 * 128 (since 27 * 2.5 = 67.5, I do not think it works, but I have no idea how to handle double in this case), the actual result done by MATLAB is 66 * 128.
How to derive the 66 in this case?
I tried to change the scale factor to 2, and the result is 53 * 128, and I expect it to be 54 * 128 since 27 * 2 = 54.
load mri
M1 = D(:,64,:,:);
M2 = reshape(M1,[128 27]);
T0 = maketform('affine',[0 -2.5; 1 0; 0 0]);
res = imtransform(M2,T0,'cubic')
size(res) // 66 * 128
A matrix with 27 elements has coordinates going from 0 through 26 (these are the coordinates used by imtransform). After scaling by 2.5, these coordinates go from 0 through 26*2.5 = 65. To hold x-coordinates from 0 through 65 you need 66 elements.
Related
Using matlab, I want to apply transform contain of rotate and translate to 2d points.
for example my points are:
points.x=[1 5 7 100 52];
points.y=[42 96 71 3 17];
points.angle=[2 6 7 9 4];
the value of rotate is:30 degree
the value of x_translate is 5.
the value of y_translate is 54.
can any body help me to write matlab code for apply this transform to my points and calculate new coordinate of points after transform?
I don't know what you mean by points.angle since the angle of the points with respect to origin (in a trigonometric sense) is already defined by atand2(y,x)
Here is the code:
clear;clc
oldCoord = [1 5 7 100 52;42 96 71 3 17];
newCoord = zeros(size(oldCoord));
theta = 30 * pi/180;
T = #(theta) [cos(theta), -sin(theta); sin(theta) , cos(theta)];
trans = [5;54];
for m = 1:size(oldCoord,2)
newCoord(:,m) = T(theta) * oldCoord(:,m) + trans;
end
Result:
oldCoord =
1 5 7 100 52
42 96 71 3 17
newCoord =
-15.1340 -38.6699 -24.4378 90.1025 41.5333
90.8731 139.6384 118.9878 106.5981 94.7224
I am struggling with the concepts behind plotting a surface polar plot.
I am trying to plot the values measured by a sensor at a combination of different angles over a hemisphere.
I have an array containing the following information:
A(:,1) = azimuth values from 0 to 360º
A(:,2) = zenith values from 0 to 90º
A(:,3) = values measured at the combination of angles of A(:,1) and A(:,2)
For example, here is a snippet:
0 15 0.489502132167206
0 30 0.452957556748497
0 45 0.468147850273115
0 60 0.471115818950192
0 65 0.352532182508945
30 15 0.424997863795610
30 30 0.477814980942155
30 45 0.383999653859467
30 60 0.509625464595446
30 75 0.440940431784788
60 15 0.445028058361392
60 30 0.522388502880219
60 45 0.428092266657885
60 60 0.429315072676194
60 75 0.358172892912138
90 15 0.493704001125912
90 30 0.508762762699997
90 45 0.450598496609200
90 58 0.468523071441297
120 15 0.501619699042408
120 30 0.561755273071577
120 45 0.489660355057938
120 60 0.475478615354648
120 75 0.482572226928475
150 15 0.423716506205776
150 30 0.426735372570756
150 45 0.448548968227972
150 60 0.478055144126694
150 75 0.437389584937356
To clarify, here is a piece of code that shows the measurement points on a polar plot.
th = A(:,1)*pi/180
polar(th,A(:,2))
view([180 90])
This gives me the following plot:
I would like now to plot the same thing, but instead of the points, use the values of these points stored in A(:,3). Then, I would like to interpolate the data to get a colored surface.
After some research, I found that I need to interpolate my values over a grid, then translate to Cartesian coordinates. From there I do not know how to proceed. Could someone point me in the right direction?
I have trouble getting the concept of the interpolation, but this is what I have attempted:
x1 = linspace(0,2*pi,100)
x2 = linspace(0,90,100)
[XX,YY] = meshgrid(x1,x2)
[x,y] = pol2cart(th,A(:,2))
gr=griddata(x,y,A(:,3),XX,YY,'linear')
With this piece of code, your example data points are converted into cartesian coords, and then plotted as "lines". The two tips of a line are one data point and the origin.
az = bsxfun(#times, A(:,1), pi/180);
el = bsxfun(#times, A(:,2), pi/180);
r = A(:,3);
[x,y,z] = sph2cart(az,el,r);
cx = 0; % center of the sphere
cy = 0;
cz = 0;
X = [repmat(cx,1,length(x));x'];
Y = [repmat(cy,1,length(y));y'];
Z = [repmat(cz,1,length(z));z'];
Still thinking how to interpolate the data so you can draw a sphere. See my comments to your question.
I have a matrix of measured angles between M planes
0 52 77 79
52 0 10 14
77 10 0 3
79 14 3 0
I have a list of known angles between planes, which is an N-by-N matrix which I name rho. Here's is a subset of it (it's too large to display):
0 51 68 75 78 81 82
51 0 17 24 28 30 32
68 17 0 7 11 13 15
75 24 7 0 4 6 8
78 28 11 4 0 2 4
81 30 13 6 2 0 2
82 32 15 8 4 2 0
My mission is to find the set of M planes whose angles in rho are nearest to the measured angles.
For example, the measured angles for the planes shown above are relatively close to the known angles between planes 1, 2, 4 and 6.
Put differently, I need to find a set of points in a distance matrix (which uses cosine-related distances) which matches a set of distances I measured. This can also be thought of as matching a pattern to a mold.
In my problem, I have M=5 and N=415.
I really tried to get my head around it but have run out of time. So currently I'm using the simplest method: iterating over every possible combination of 3 planes but this is slow and currently written only for M=3. I then return a list of matching planes sorted by a matching score:
function [scores] = which_zones(rho, angles)
N = size(rho,1);
scores = zeros(N^3, 4);
index = 1;
for i=1:N-2
for j=(i+1):N-1
for k=(j+1):N
found_angles = [rho(i,j) rho(i,k) rho(j,k)];
score = sqrt(sum((found_angles-angles).^2));
scores(index,:)=[score i j k];
index = index + 1;
end
end;
end
scores=scores(1:(index-1),:); % was too lazy to pre-calculate #
scores=sortrows(scores, 1);
end
I have a feeling pdist2 might help but not sure how. I would appreciate any help in figuring this out.
There is http://www.mathworks.nl/help/matlab/ref/dsearchn.html for closest point search, but that requires same dimensionality. I think you have to bruteforce find it anyway because it's just a special problem.
Here's a way to bruteforce iterate over all unique combinations of the second matrix and calculate the score, after that you can find the one with the minimum score.
A=[ 0 52 77 79;
52 0 10 14;
77 10 0 3;
79 14 3 0];
B=[ 0 51 68 75 78 81 82;
51 0 17 24 28 30 32;
68 17 0 7 11 13 15;
75 24 7 0 4 6 8;
78 28 11 4 0 2 4;
81 30 13 6 2 0 2;
82 32 15 8 4 2 0];
M = size(A,1);
N = size(B,1);
% find all unique permutations of `1:M`
idx = nchoosek(1:N,M);
K = size(idx,1); % number of combinations = valid candidates for matching A
score = NaN(K,1);
idx_triu = triu(true(M,M),1);
Atriu = A(idx_triu);
for ii=1:K
partB = B(idx(ii,:),idx(ii,:));
partB_triu = partB(idx_triu);
score = norm(Atriu-partB_triu,2);
end
[~, best_match_idx] = min(score);
best_match = idx(best_match_idx,:);
The solution of your example actually is [1 2 3 4], so the upperleft part of B and not [1 2 4 6].
This would theoretically solve your problem, and I don't know how to make this algorithm any faster. But it will still be slow for large numbers. For example for your case of M=5 and N=415, there are 100 128 170 583 combinations of B which are a possible solution; just generating the selector indices is impossible in 32-bit because you can't address them all.
I think the real optimization here lies in cutting away some of the planes in the NxN matrix in a preceding filtering part.
I know how to check an 8-neighbourhood in matlab (i.e; nlfilter). But, I want to assign the value which is more repeated to the center value. So, say for instance that I have the following values in the 8-neighbourhood:
2-values = 56
3-values = 64
1-value = 70
1-value = 87
1-value = 65
In this case we would assign 64 to the center pixel.
How can we do that?
Thanks.
I think you want either the mode or the histc function.
M=mode(X) for vector X computes M as the sample mode, or most
frequently
occurring value in X.
Example with your data:
x = [56 56 64 64 64 70 87 65];
mode(x)
ans =
64
But this will only get you the most frequently occurring value.
If you want the count of each unique item in the array, you could do,
unqx = unique(x);
unqx =
56 64 65 70 87
valueCount = histc(x, unqx)
ans =
2 3 1 1 1
You could then sort this and take the first N values
valueCount = sort(valueCount, 'descend');
% Use unqx(valueCount(1:N))
I want to calculate the sum of the elements surrounding a given element in a matrix. So far, I have written these lines of code:
for i=1:m,
rij(1:n)=0
for j=1:n,
alive = tijdelijk(i-1,j)+tijdelijk(i+1,j)+tijdelijk(i-1,j-1)+tijdelijk(i+1,j-1)+tijdelijk(i,j+1)+tijdelijk(i,j-1)+tijdelijk(i-1,j+1)+tijdelijk(i+1,j+1)
This results in an error because, for example, i-1 becomes zero for i=1. Anyone got an idea how to do this without getting this error?
You can sum the elements via filtering. conv2 can be used for this manner.
Let me give an example. I create a sample matrix
>> A = reshape(1:20, 4, 5)
A =
1 5 9 13 17
2 6 10 14 18
3 7 11 15 19
4 8 12 16 20
Then, I create a filter. The filter is like a mask where you put the center on the current cell and the locations corresponding to the 1's on the filter are summed. For eight-connected neighbor case, the filter should be as follows:
>> B = [1 1 1; 1 0 1; 1 1 1]
B =
1 1 1
1 0 1
1 1 1
Then, you simply convolve the matrix with this small matrix.
>> conv2(A, B, 'same')
ans =
13 28 48 68 45
22 48 80 112 78
27 56 88 120 83
18 37 57 77 50
If you want four-connected neighbors, you can make the corners of your filter 0. Similarly, you can design any filter for your purpose, such as for averaging all neighbors instead of summing them.
For details, please see the convolution article in Wikipedia.
Two possibilities : change the limits of the loops to i=k:(m-k) and j=k:(n-k) or use blkproc
ex :
compute the 2-D DCT of each 8-by-8 block
I = imread('cameraman.tif');
fun = #dct2;
J = blkproc(I,[8 8],fun);
imagesc(J), colormap(hot)
There are lots of things you can do at the edges. Which you do depends very specifically on your problem and is different from usage case to usage case. Typical things to do:
If (i-1) or (i+1) is out of range, then just ignore that element. This is equivalent to zero padding the matrix with zeros around the outside and adjusting the loop limits accordingly
Wrap around the edges. In other words, for an MxN matrix, if (i-1) takes you to 0 then instead of taking element (i-1, j) = (0, j) you take element (M, j).
Since your code mentions "your teacher" I'd guess that you can ask what should happen at the edges (or working it out in a sensible manner may well be part of the task!!).