4 ***********************
8 In this tutorial you will learn how to:
10 * Use the OpenCV function :hough_circles:`HoughCircles <>` to detect circles in an image.
15 Hough Circle Transform
16 ------------------------
18 * The Hough Circle Transform works in a *roughly* analogous way to the Hough Line Transform explained in the previous tutorial.
19 * In the line detection case, a line was defined by two parameters :math:`(r, \theta)`. In the circle case, we need three parameters to define a circle:
23 C : ( x_{center}, y_{center}, r )
25 where :math:`(x_{center}, y_{center})` define the center position (gree point) and :math:`r` is the radius, which allows us to completely define a circle, as it can be seen below:
27 .. image:: images/Hough_Circle_Tutorial_Theory_0.jpg
28 :alt: Result of detecting circles with Hough Transform
31 * For sake of efficiency, OpenCV implements a detection method slightly trickier than the standard Hough Transform: *The Hough gradient method*. For more details, please check the book *Learning OpenCV* or your favorite Computer Vision bibliography
36 #. **What does this program do?**
38 * Loads an image and blur it to reduce the noise
39 * Applies the *Hough Circle Transform* to the blurred image .
40 * Display the detected circle in a window.
42 .. |TutorialHoughCirclesSimpleDownload| replace:: here
43 .. _TutorialHoughCirclesSimpleDownload: http://code.opencv.org/projects/opencv/repository/revisions/master/raw/samples/cpp/houghcircles.cpp
44 .. |TutorialHoughCirclesFancyDownload| replace:: here
45 .. _TutorialHoughCirclesFancyDownload: http://code.opencv.org/projects/opencv/repository/revisions/master/raw/samples/cpp/tutorial_code/ImgTrans/HoughCircle_Demo.cpp
47 #. The sample code that we will explain can be downloaded from |TutorialHoughCirclesSimpleDownload|_. A slightly fancier version (which shows both Hough standard and probabilistic with trackbars for changing the threshold values) can be found |TutorialHoughCirclesFancyDownload|_.
51 #include "opencv2/highgui/highgui.hpp"
52 #include "opencv2/imgproc/imgproc.hpp"
59 int main(int argc, char** argv)
64 src = imread( argv[1], 1 );
69 /// Convert it to gray
70 cvtColor( src, src_gray, CV_BGR2GRAY );
72 /// Reduce the noise so we avoid false circle detection
73 GaussianBlur( src_gray, src_gray, Size(9, 9), 2, 2 );
75 vector<Vec3f> circles;
77 /// Apply the Hough Transform to find the circles
78 HoughCircles( src_gray, circles, CV_HOUGH_GRADIENT, 1, src_gray.rows/8, 200, 100, 0, 0 );
80 /// Draw the circles detected
81 for( size_t i = 0; i < circles.size(); i++ )
83 Point center(cvRound(circles[i][0]), cvRound(circles[i][1]));
84 int radius = cvRound(circles[i][2]);
86 circle( src, center, 3, Scalar(0,255,0), -1, 8, 0 );
88 circle( src, center, radius, Scalar(0,0,255), 3, 8, 0 );
92 namedWindow( "Hough Circle Transform Demo", CV_WINDOW_AUTOSIZE );
93 imshow( "Hough Circle Transform Demo", src );
108 src = imread( argv[1], 1 );
113 #. Convert it to grayscale:
117 cvtColor( src, src_gray, CV_BGR2GRAY );
119 #. Apply a Gaussian blur to reduce noise and avoid false circle detection:
123 GaussianBlur( src_gray, src_gray, Size(9, 9), 2, 2 );
125 #. Proceed to apply Hough Circle Transform:
129 vector<Vec3f> circles;
131 HoughCircles( src_gray, circles, CV_HOUGH_GRADIENT, 1, src_gray.rows/8, 200, 100, 0, 0 );
135 * *src_gray*: Input image (grayscale)
136 * *circles*: A vector that stores sets of 3 values: :math:`x_{c}, y_{c}, r` for each detected circle.
137 * *CV_HOUGH_GRADIENT*: Define the detection method. Currently this is the only one available in OpenCV
138 * *dp = 1*: The inverse ratio of resolution
139 * *min_dist = src_gray.rows/8*: Minimum distance between detected centers
140 * *param_1 = 200*: Upper threshold for the internal Canny edge detector
141 * *param_2* = 100*: Threshold for center detection.
142 * *min_radius = 0*: Minimum radio to be detected. If unknown, put zero as default.
143 * *max_radius = 0*: Maximum radius to be detected. If unknown, put zero as default
145 #. Draw the detected circles:
149 for( size_t i = 0; i < circles.size(); i++ )
151 Point center(cvRound(circles[i][0]), cvRound(circles[i][1]));
152 int radius = cvRound(circles[i][2]);
154 circle( src, center, 3, Scalar(0,255,0), -1, 8, 0 );
156 circle( src, center, radius, Scalar(0,0,255), 3, 8, 0 );
159 You can see that we will draw the circle(s) on red and the center(s) with a small green dot
161 #. Display the detected circle(s):
165 namedWindow( "Hough Circle Transform Demo", CV_WINDOW_AUTOSIZE );
166 imshow( "Hough Circle Transform Demo", src );
168 #. Wait for the user to exit the program
178 The result of running the code above with a test image is shown below:
180 .. image:: images/Hough_Circle_Tutorial_Result.jpg
181 :alt: Result of detecting circles with Hough Transform