Image Processing and Hough Transform for Circle Detection

This project focuses on implementing circle detection in an image using the Hough transform. The core of the solution involves applying the circular Hough transform to detect circles based on their intensity values, with a specific focus on a sample image containing 17 circular coins. The assignment details the use of a radius range ([25 60]) to accurately identify the circles and their centers. The code utilizes MATLAB scripts (CircleDetection.m and createcircle.m) and a test script (test.m) to process the image. The output includes the number of detected circles (k=17), the calculated center coordinates and the radius of the circles. The project visualizes the results by plotting the estimated centers and perimeters on the original image, along with an accumulated array image display. The project demonstrates the practical application of the Hough transform in image processing, providing insights into the detection of circular objects.

Circle detection of an image using Hough transform:
The method of circle detection in an image is based on the circular Hough transform of the
image. The circles in an image has different intensity values than the background. The Hough
transform uses the radius range in unit if pixels to detect the circles within the given range.
The accumulation array of pixel values after applying Hough transform is displayed here.
The chosen image sample.png has 17 circular coins in total. The circle radius pixel range
which accurately detects 17 circles and their centres is [25 60]. Additionally the estimated
centres of the circle and their predicted perimeter are plotted in the same raw image.
At first Circledetection.m and createcircle.m are taken into same directory for execution and
then test.m is executed.
Output:
k = 17
centrecirc =
64.0092 339.1048
65.4000 564.8232
87.8493 214.6536
186.3726 506.6008
190.0934 93.8637
310.6884 259.8257
336.4414 457.1281
360.9927 69.1053
371.4443 271.9019
501.8613 92.0709
506.8821 289.5313
564.2064 533.4465
564.4950 221.5493
687.7796 117.2623
703.7101 486.6580
799.6594 184.1339
805.7633 358.1163
radiuscirc

radiuscirc =
30
30
28
28
31
28
30
28
28
27
30
31
27
28
31
31
29
Accumulated array image display:

Accumulated Array after application of the Circular Hough Transform
100 200 300 400 500 600 700 800
100
200
300
400
500
600
Circles detected in the grayscale image:
Detected circles in the given image (position of centres and their radius)
100 200 300 400 500 600 700 800
100
200
300
400
500
600