From the given matrices A and B I need to compute a new matrix C. Matrix A represents image pixels and C is a horizontally shifted version of A. The tricky part: this shift is defined per pixel by the values in the disparity matrix B. For exampe: While pixel (1,1) needs to be shifted 0.1 units to the right, pixel (1,2) needs to be shifted 0.5 units to the left.
I implemented this as backward-mapping, where for each pixel in C I compute the required source position in A (which is simply my current pixel's location minus the corresponding offset in B). Since non-integer shifts are allowed, I need to interpolate the new pixel value.
Doing this in Matlab, of course, takes quite some time as images get larger. Is there any built-in function I can utilize for this task?
I assume that matrix A is an image, which means that the pixels are regularly spaced, which means you can use INTERP2. I also assume that you calculate for each pixel individually the interpolated value from A. You can, however, perform the lookup in one step, which will be quite a bit faster.
Say A is a 100x100 image, and B is a 10000-by-2 array with [shiftUpDown,shiftLeftRight] for each pixel. Then you'd calculate C this way:
%# create coordinate grid for image A
[xx,yy] = ndgrid(1:100,1:100);
%# linearize the arrays, and add the offsets
xx = xx(:);
yy = yy(:);
xxShifted = xx + B(:,1);
yyShifted = yy + B(:,2);
%# preassign C to the right size and interpolate
C = A;
C(:) = interp2(xx,yy,A(:),xxShifted,yyShifted);
The function interp2 interpolates values on a regularly spaced grid, such as a bitmap image. If your pixels didn't lie on a regular grid, then you would use griddata.
Related
I have two sets of 3D images (they come in form of 2D stacks). Image A is 10 micron, with size: 1000 x 1024 x 1017, while image B is 5 micron, with size: 2004 x 2048 x 2036. I like to make some computations on randomly chosen set of the 2D slices of A, and then compare this to the same set of slices for B. However, since B has twice the number of slices for each slice of A, will it be sensible to compare two slices of B to each of A? If so, how do i determine exactly which of the two slices of B make up a slice of A?
While contemplating on this, i also thought of blowing up A by 2 using imresize function for each 2D slice that i chose for the computation. Will it be okay to compare this new B with the A, considering that i have completely ignored what happens with the z-coordinate?
Thank you.
As you mentioned this is microCT, I am assuming that both images are different size resolution of the same object. This means that pixels have specific spatial location, not only value, therefore for this case, there are no pixels (assuming a pixel is a infinitesimally small dot in the center of the cube) that match in both images.
So, lets assume that in image A, the locations of the pixel centers are their indices (1,1,1), (1,1,2) etc. This means that the image starts (pixel boundaries) at "world value" 0.5, and ends at size(imgA)+0.5
Now, first lets transform the desired pixel coordinates for interpolation to this range. imgB pixel centers are then in locations (ind-0.5)*size(imgA)/size(imgB)+0.5.
Example: Assume
size(imgA,1)=3; size(imgB,1)=4;
therefore the pixels in imgA are at x location 1:3. The pixels on imgB are, using the above formula, in [0.8750 1.6250 2.3750 3.1250].
Note how the first pixel is 0.375 from 0.5 (our image border) and the next pixel is at 0.75 (double 0.375*2).
We scaled a higher resolution image to the same "real world" coordinates.
Now to the real stuff.
We need to create the desired coordinates in the reference (A) image. For that, we do:
[y, x, z]=...
ndgrid((1:size(imgB,1)-0.5)*size(imgA,1)/size(imgB,1)+0.5),...
(1:size(imgB,2)-0.5)*size(imgA,2)/size(imgB,2)+0.5),...
(1:size(imgB,3)-0.5)*size(imgA,3)/size(imgB,3)+0.5);
Now these 3 have the coordinates we want. Caution! each of these are size(imgB) !!! You need to have the RAM 5*size(imgB) in total to work with this.
Now we can interpolate
imAinB=interp3(imgA,x,y,z,'linear'); % Or nearest
It seems to be that your function is imresize3. You can change one volume to the other's dimentions with:
B = imresize3(V,[numrows numcols numplanes])
You can also explore interpolation methods.
I have a matrix with grey values between 0 and 1. For every entry in the matrix, there are certain polar coordinates that indicate the position of the grey values. I already have either Theta and Rho values (polar) ,both in separate 512×960 matrices. And grayscale values (in a matrix called C) for every Theta and Rho combination. I have the same for X and Y, as I just use pol2cart for the transformation. The problem is that I cannot directly plot these values, as they do not yet fit in the 'bins' of the new matrix.
What I want: to put the grey values in a square matrix of size 1024×1024. I cannot do this directly, because the polar coordinates fall in between the grid of this matrix. Therefore, we now use interpolation, but this is extremely time consuming and has to be done separately for every dataset, although the transformation from the original matrices to this final one will always be the same. Therefore, I'd like to solve this matrix once (either analytically or numerically) and use a matrix multiplication or something similar to apply the manipulation efficiently in every cycle of the code.
One example of what one of these transformations could look like this:
The zeros in the first matrix are the grid, and the value 1 (in between the grid) is the grey value that falls in between four grid points, then I'd like to transform to the second matrix (don't mind the visual spacing between the points).
For every dataset, I have hundreds of these matrices, so I would like to make the code more efficient.
Background: I'm using TriScatteredInterp now for the interpolation. We tried scatteredInterpolant as well, but that is slower. I also posted a related question, but decided to split the two possible solutions, because the solution I ask for here is also applicable to non-MATLAB code and will probably be faster and makes for a smoother (no continuous popping up of figures) execution of the code.
Using the image processing toolbox
Images work a bit differently than the data you have. However, it's fairly straightforward to map one representation into the other.
There is only one problem I see: wrapping. Obviously, θ = 2π = 0, but MATLAB does not know that. AFAIK, there is no easy way to tell MATLAB that.
Why does this matter? Well, simply put, inter-pixel interpolation uses information from the nearest N neighbors to find intermediate colors, with N depending on the interpolation kernel. When doing this somewhere in the middle of the image there is no problem, but at the edges MATLAB has to know that the left edge equals the right edge. This is not standard image processing, and I'm not aware of any function that is capable of this.
Implementation
Now, when disregarding the wrapping problem, this is one way to do it:
function resize_polar()
%% ORIGINAL IMAGE
% ==========================================================================
% Some random greyscale data
C = double(rgb2gray(imread('stars.png')))/255;
% Your current size, and desired size
sz_x = size(C,2); new_sz_x = 1024;
sz_y = size(C,1); new_sz_y = 1024;
% Ranges for teat and rho;
% replace with your actual values
rho_start = 0; theta_start = 0;
rho_end = 10; theta_end = 2*pi;
% Generate regularly spaced grid;
theta = linspace(theta_start, theta_end, sz_x);
rho = linspace(rho_start, rho_end, sz_y);
[theta, rho] = meshgrid(theta,rho);
% Make plot of generated data
plot_polar(theta, rho, C, 'Original image');
% Resize data
[theta,rho,C] = resize_polar_data(theta, rho, C, [new_sz_y new_sz_x]);
% Make plot of generated data
plot_polar(theta, rho, C, 'Rescaled image');
end
function [theta,rho,data] = resize_polar_data(theta,rho,data, new_dims)
% Create fake RGB image cube
IMG = cat(3, theta,rho,data);
% Rescale as if theta and rho are RG color data in the RGB
% image cube
IMG = imresize(IMG, new_dims, 'nearest');
% Split up the data again
theta = IMG(:,:,1);
rho = IMG(:,:,2);
data = IMG(:,:,3);
end
function plot_polar(theta, rho, data, label)
[X,Y] = pol2cart(theta, rho);
figure('renderer', 'opengl')
clf, hold on
surf(X,Y,zeros(size(X)), data, ...
'edgecolor', 'none');
colormap gray
title(label);
end
The images used and plotted:
Le awesomely-drawn 512×960 PNG image
Now, the two look the same (couldn't really come up with a better-suited image), so you'll have to believe me that the 512×960 has indeed been rescaled to 1024×1024, with nearest-neighbor interpolation.
Here are some timings for the actual imresize() operation for some simple kernels:
nearest : 0.008511 seconds.
bilinear: 0.019651 seconds.
bicubic : 0.025390 seconds. <-- default kernel
But this depends strongly on your hardware; I believe imresize offloads a lot of work to the GPU, so if you have a crappy one, it'll be slower.
Wrapping
If the wrapping problem is really important to you, you can modify the function above to do the following:
first, rescale the image with imresize() like before
horizontally concatenate the second half of the grayscale data and the first half. Meaning, you swap the first and second halves to make the left and right edges (0 and 2π) touch in the middle.
rescale this intermediate image with imresize()
Extract the central vertical strip of the rescaled intermediate image
split that up in two equal-width strips
and replace the edge strips of the output image with the two strips you just created
Now, this is kind of a brute force approach: you are re-scaling an image twice, and most of the pixels of the second image round will be discarded. If performance is a problem, you can of course apply the rescale to only the central strip of that intermediate image. But, well, that will be a bit more complicated.
I want to build a contourf plot of a certain aspect in my Plate. The plate is divided in triangle elements, which I have the coordinates (x,y) of each knot of the triangle.
So, How can I make a meshgrid for my knots so I can make my contourf plot?? I have the coordinates of everything and have the value of my function Z in each knot. (I'm a beginner in Matlab, sorry for this "basic" question)
If your goal is just to visualise the triangles then there is another way that's probably simpler and more robust (see the end of this post).
If you definitely need to generate contours then you will need to interpolate your triangular mesh over a grid. You can use the scatteredInterpolant class for this (documentation here). It takes the X and Y arguments or your triangular vertices (knots), as well as the Z values for each one and creates a 'function' that you can use to evaluate other points. Then you create a grid, interpolate your triangular mesh over the grid and you can use the results for the countour plot.
The inputs to the scatteredInterpolanthave to be linear column vectors, so you will probably need to reshape them using the(:)`notation.
So let's assume you have triangular data like this
X = [1 4; 8 9];
Y = [2 3; 4 5];
Z = [0.3 42; 16 8];
you would work out the upper and lower limits of your range first
xlimits = minmax(X(:));
ylimits = minmax(Y(:));
where the (:) notation serves to line up all the elements of X in a single column.
Then you can create a meshgrid that spans that range. You need to decide how fine that grid should be.
spacing = 1;
xqlinear = xlimits(1):spacing:xlimits(2);
yqlinear = ylimits(1):spacing:ylimits(2);
where linspace makes a vector of values starting at the first one (xlimits(1)) and ending at the third one (xlimits(2)) and separated by spacing. Experiment with this and look at the results, you'll see how it works.
These two vectors specify the grid positions in each dimension. To make an actual meshgrid-style grid you then call meshgrid on them
[XQ, YQ] = meshgrid(xqlinear, yqlinear);
this will produce two matrices of points. XQ holds the x-coordinates of every points in the grid, arranged in the same grid. YQ holds the y-coordinates. The two need to go together. Again experiment with this and look at the results, you'll see how it works.
Then you can put them all together into the interpolation:
F = scatteredInterpolant(X(:), Y(:), Z(:));
ZQ = F(XQ, YQ);
to get the interpolated values ZQ at each of your grid points. You can then send those data to contourf
contourf(XQ, YQ, ZQ);
If the contour is too blocky you will probably need to make the spacing value smaller, which will create more points in your interpolant. If you have lots of data this might cause memory issues, so be aware of that.
If your goal is just to view the triangular mesh then you might find trimesh does what you want or, depending on how your data is already represented, scatter. These will both produce 3D plots with wireframes or point clouds though so if you need contours the interpolation is the way to go.
So i have this code to obtain radial gravity on Earth in function of the latitude:
G=6.6e-11;
M=5.976e24;
N=1000;
r=6371000;
w=2*pi/(24*3600);
for i=1:1:360
Theta=i*pi/180;
x(i)=i;
Vi(i)=-G*M/(r*r);
Phii(i)=r*w*w*sin(Theta)*sin(Theta);
gr(i)=Vi(i)+Phii(i);
end
plot(x,gr)
And it runs well. I want to make a graph of a circle made of a border of points (representing angle (i)) that change colour according to the value of gr(I want to set ranges of values of gr so that if the value obtained falls in a specific category, the point will have a specific colour).
I'm really new to MATLAB. Is there any possible way to make this?
Thanks in advance.
Here is the basic algorithm that I would do:
Determine how many colours you want to represent in your plot.
Create a colour map that has this many points for what you want to compute.
Determine a linearly increasing vector that varies from the minimum value of gr to the maximum value of gr with as many points as you have determined in Step #2
For each point in gr:
a. Determine which point yields the closest distance of this point to the vector in Step #3
b. Use this to index which colour you want.
c. Convert your angle into Cartesian co-ordinates, then plot this point with the colour found in Step 4b.
Let's tackle each point in detail.
Step #1 - Determine how many colours you want
This is pretty simple. Just determine how many colours you want. For now, let's assume that you want 20 colours, so:
num_colours = 20;
Step #2 - Create a colour map
What we can do is create a 20 x 3 matrix where each row determines a RGB tuple that denotes the amount of red, green and blue that each colour will occupy. MATLAB has built-in colour maps that will help you facilitate this. Here are all of the available colour maps that MATLAB has:
Each colour map has a special variable where you can provide it an integer number, and it'll return this 2D matrix of as many rows as the number you have provided. Each row gives you an RGB triplet which denotes the proportion of red, green and blue respectively. This matrix varies from the beginning of the colour map (top row) to the end (bottom row). All you have to do is use any name seen in the figure I've shown you above to create a colour map of that type. For example, if you wanted to get a bones colour map of 15 points, simply do:
colour_map = bones(15);
If you wanted to get a jet colour map of 25 points, simply do:
colour_map = jet(25);
.... you get the idea right? I like hsv so let's use the HSV colour map. You can use any colour map you want, but let's just stick with HSV for the sake of this example.
As such:
colour_map = hsv(num_colours);
Step #3 - Get that linearly increasing vector
You want certain colours to map into certain ranges, which is why this step is important. Given a value in gr, we want to figure out which colour we want to choose, and all you have to do is determine which value in gr is the closest to a value in this vector in Step #3. Therefore, you can use linspace to do this for you:
bin_vector = linspace(min(gr), max(gr), num_colours);
This will create a num_colours 1D array where the beginning of this array starts at the minimum value of gr and varies up to the maximum value of gr and each value is equally spaced such that we generate a num_colours array.
Step #4 - Bring it all home
Now, for each point in gr, we need to figure out which point is the closest to that vector in Step #3, we then use this to figure out the colour we want, then we need to convert our angle into Cartesian co-ordinates, then plot this point.
For illustration purposes, I'm going to assume your radius is 1. You can figure out how to get the x and y co-ordinates by simply doing cos(theta) and sin(theta), where theta is the angle you are examining. Since your gr array has 360 slots, I'm going to assume a resolution of 1 degree per slot. Therefore, you can easily do this in a for loop. Make sure you use hold on because we are going to call plot multiple times, and we don't want to overwrite the plot each time we call plot. You want all of the points to stay in the plot.
Without further ado:
figure; %// Create blank figure
hold on; %// Remember all points
%// For each point in our array...
for idx = 1 : 360
%// Find the closest slot between gr and our vector in Step #3
[~,min_idx] = min(abs(gr(idx) - bin_vector));
%// Grab this colour
clr = colour_map(min_idx,:);
%// Plot the point with this colour
plot(cosd(idx), sind(idx), '.', 'Color', clr, 'MarkerSize', 10);
end
Take notice that cosd and sind take in degrees as the input argument while cos and sin take in radians. Also, take note that I also changed the size of the point so that it's bigger. With the above logic, and your array in gr, this is what I get:
If you want the radius to get larger, all you have to do is multiply each cosd and sind term with your radius. Therefore, you can do something like this:
radius = 2;
for idx = 1 : 360
... %// Insert colour code here
...
...
%// Now plot
plot(radius*cosd(idx), radius*sind(idx), '.', 'Color', clr, 'MarkerSize', 10);
end
Just leave the code the same, but for the plot command, just multiply each x and y value by the radius.
Minor note in efficiency
The way you're calculating your gr array is using an inefficient for loop. There are some situations (like mine above) where you need to use a for loop, but for simple computations there is no need. It's better if you vectorize its creation. Therefore, you can get rid of the for loop to calculate your gr array like so:
x = 1 : 360;
Theta = x*pi/180;
Phii = r*w*w*sin(Theta).*sin(Theta);
Vi = -G*M/(r*r);
gr = Vi + Phii;
x is simply a vector going from 1 to 360, and that's done in the first line. Also, Vi is just an array which contains a single value and if you know how operations work between a scalar and an array, you can just do an addition with this single value and it'll add every value in your array by this much. As such, there's no need to create an array for Vi. Also, take a look at how I calculated Phii. I'm using element-by-element operations as Theta is now an array. You want to create an array Phii that takes corresponding values of Theta, and applies that formula to each value in Theta to produce Phii.
Hope this helps. Good luck!
I want to calculate the magnetic field from a given image using biot savarts law. For example if I have a picture of a triangle, I say that this triangle forms a closed wire carrying current. Using image derivatives I can get the co-ordinates and direction of the current (normals included). I am struggling implementing this...need a bit of help with logic too. Here is what I have:
Img = imread('littletriangle.bmp');
Img = Img(:,:,1);
Img = double(Img);
[x,y] = size(Img);
[Ix, Iy] = gradient(Img);
biot savart equation is:
b = mu/4*pi sum(Idl x rn / r^2)
where mu/4pi is const, I is current magnitude, rn distance unit vector between a pixel and current, r^2 is the squared magnitude of the displacement between a pixel and the current.
So just to start off, I read the image in, turn it into a binary and then take the image gradient. This gives me the location and orientation of the 'current'. I now need to calculate the magnetic field from this 'current' at every pixel in the image. I am only interested in getting the magnetic field in the x-y plane. anything just to start me off would be brilliant!
For wire
B = mu * I /(2*pi*r)
B is vector and has. Direction is perpendicular on line between wire an point of interest. Fastest way to rotate vector by 90° is just swapping (x.y) so it becomes (y,x) vector
What about current? If you deal whit current then current is homogenous inside of wire (can be triangle) and I in upper direction is just normalized I per point and per Whole I.
So how to do this?
Get current per pixel (current / number of pixel in shape)
For each point calculate B using (r calculated form protagora) as sum of all other mini wires expressed as pixel using upper equation. (B is vector and has also direction, so keep track of B as (x,y) )
having picture of 100*100 will yield (100*100)*(100*100) calculations of B equation or something less if you will not calculate filed from empty space.
B is at the end instead of just mu * I /(2*pi*r) sum of all wire and I becomes dI
You do not need to apply any derivatives, just integration (sum)