Scribd
Upload
30 views

Hough Examples - Algorithm

The document contains Matlab code and images from Grant Latham demonstrating the use of the Hough transform to detect various shapes including lines, circles, and parabolas. For lines, it shows detecting lines in an image using edge detection and the Hough transform, then drawing the detected lines back on the original image. For circles, it shows detecting circles in noisy coin images and in video frames of a moving object. For parabolas, it introduces the equation for a parabola and pseudocode for detecting parabolas using a 3D Hough accumulator array to search for the parameters defining each parabola shape.

Uploaded by

ashenafiay2015
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
30 views

Hough Examples - Algorithm

The document contains Matlab code and images from Grant Latham demonstrating the use of the Hough transform to detect various shapes including lines, circles, and parabolas. For lines, it shows detecting lines in an image using edge detection and the Hough transform, then drawing the detected lines back on the original image. For circles, it shows detecting circles in noisy coin images and in video frames of a moving object. For parabolas, it introduces the equation for a parabola and pseudocode for detecting parabolas using a 3D Hough accumulator array to search for the parameters defining each parabola shape.

Uploaded by

ashenafiay2015
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Example 1 ‐ lines

clear all
close all

I = imread('hallway.jpg');
imshow(I,[]);

E = edge(I, 'canny');
figure, imshow(E);

[H, theta, rho] = hough(E);


peaks = houghpeaks(H, 10);

% Draw each line


figure, imshow(I,[]); image “hallway.jpg”
for i=1:size(peaks,1)
% Extract rho, theta for this line
r = rho(peaks(i,1));
t = theta(peaks(i,2));

% Equation of the line is r = x cos(t) + y sin(t), or


% y = (r - x*cos(t))/sin(t)
x0 = 1;
y0 = (r - x0*cosd(t))/sind(t);
x1 = size(I,2);
y1 = (r - x1*cosd(t))/sind(t);

line([x0 x1], [y0 y1], 'Color', 'r');


end
Example 1 (cont)

clear all
close all

I = imread('hallway.jpg');
imshow(I,[]);

E = edge(I, 'canny');
figure, imshow(E);
image “hallway.jpg”
[H, theta, rho] = hough(E);
peaks = houghpeaks(H, 10);

% Use Matlab's Hough lines


lines = houghlines(E, theta, rho, peaks);

figure, imshow(I);
for k = 1:length(lines)
xy = [lines(k).point1; lines(k).point2];
line(xy(:,1),xy(:,2),'LineWidth',1.5,'Color','g');
end
Example 2 ‐ circles
% Find circles using Hough transform.
clear all
close all

I = imread('coins.png');
imshow(I,[]);

% Add noise and occlude some of the coins


I(100:120, :) = 100;
I = imnoise(I, 'gaussian', 0, 0.001);
figure, imshow(I,[]);

[centers, radii] = imfindcircles(I, ...


[20, 35], ... % Range of radii to search for
'ObjectPolarity', 'bright', ... % Circles are brighter than bkgnd
'Sensitivity', 0.9); % Higher values yield more circles

figure, imshow(I, []);


for j=1:size(centers,1)
r = radii(j);
x = centers(j,1);
y = centers(j,2);
rectangle('Position', [x-r y-r 2*r 2*r], 'EdgeColor', 'r', ...
'Curvature', [1 1], 'LineWidth', 2);
end
Example 3 – circles in video
clear all
close all

movie = VideoReader('sphero1.wmv');
images = read(movie); % get all images
nImg = size(images,4); % Number of images read
fprintf('Read in %d images from video file\n', nImg);

% Suppress warnings about low accuracy because of small radii.


warning('off', 'all');

for i=1:2:nImg
I = images(:,:,:,i); % Get next image
G = rgb2gray(I);

[centers, radii] = imfindcircles(G, [5, 20]);

imshow(I, []), title(sprintf('%d', i));


for j=1:size(centers,1)
r = radii(j);
x = centers(j,1);
y = centers(j,2);
rectangle('Position', [x-r y-r 2*r 2*r], 'EdgeColor', 'r', ...
'Curvature', [1 1], 'LineWidth', 2);
end
drawnow; Matlab code and images
end
from Grant Latham
Example 4 –parabolas
• A parabola centered at
(x0,y0) has the
equation
y  y0  a  x  x0 
2

• So we need to search
for three parameters:
x0,y0,a

• Can limit range of “a”


to search over Grenouilles.jpg

5
Pseudocode
Input a binary edge image E(x,y)
Initialize accumulator array A(i,j,k) to zeros

for all values of (x,y)


if E(x,y) == 1
for all values of a between amin and amax
a  x0
x  x0 
2
y0  y of
for all values
Compute

 
Increment A(i,j,k) where (i,j,k) corresponds to the cell
associated with x0i , y 0 j , ak
end
end
end
End

Search for peaks in A(i,j,k) – the corresponding values of


are the parameters of the detected parabolas

6
% Find parabolas in binary image I(N,M).
% A parabola is y=rx^2
clear all
close all

I = rgb2gray(imread('Grenouilles.jpg'));
imshow(I,[]);
[N,M] = size(I);

[E,thresh] = edge(I, 'canny', 0.45);


figure, imshow(E,[]);

rvals = [ 0.003, 0.004, 0.005, 0.006, 0.007, 0.008 ];


R = length(rvals); % number of sizes to try

% Fill accumulator array A(N,M,R)


A = zeros(N,M,R);
[yIndex xIndex] = find(E); % get edge points
for cnt = 1:length(xIndex)
for r=1:R
for x0 = 1:M
y0 = yIndex(cnt)-rvals(r)*(xIndex(cnt)-x0)^2;
y0 = round(y0);
if y0 < N & y0 >= 1
A(y0,x0,r) = A(y0,x0,r)+1;
end
end
end
end

figure, imshow(A(:,:,round(R/2)),[]);
title(sprintf('A slice of the accumulator array, for r=%f', ...
rvals(round(R/2)) ));
7pause
% Find the local maxima in a 3D neighborhood
Amax = imdilate(A, ones(20,20,4));

% We want those places where A = AMax and A > thresh


thresh = 95;
Apeaks = (A==Amax) & (A > thresh);

% Get indices of the peaks (the nonzero points).


% These are the indices of the array, which is treated
% like a one dimensional array.
indices1D = find(Apeaks);

% The array is actually three dimensional, not one dimensional.


% This function will calculate the equivalent indices of the three
% dimensional array.
[rowIndices,colIndices,depthIndices] = ...
ind2sub(size(Apeaks), indices1D);

figure, imshow(E,[]);
for i=1:length(rowIndices)
x0 = colIndices(i);
y0 = rowIndices(i);
r0 = rvals( depthIndices(i) );
for x=1:M
y = round(y0 + r0*(x-x0)^2);
if y<=N && y >= 1
rectangle('Position', [x y 1 1], ...
'EdgeColor', 'r');
end
end
end

You might also like