BOIECOX_ED_CVIP

boiecox_ed_cvip() - Boiecox edge detection of an input image.

SYNTAX
DESCRIPTION
REFERENCE
EXAMPLE
CREDITS

SYNTAX

out_img  = boiecox_ed_cvip( input_img, va, hys_th, thin, high_th, low_th )

Input Parameters Include:

input_img - The Input Image. Can be Single band or Multi band.

va - Variance of the Gaussian blur Kernel. It decides the size of the Gaussian kernel used. Ranges from 0.5 to 10.

hys_th - Hysteresis Threshold. It is either 0 or 1. '1' indicates Hysteresis thresholding. '0' indicates no hysteresis thresholding. Default value : 1.

thin - Morphological thinning of output.It is either 0 or 1. '1' indicates thinning operation. '0' indicates no thinning operation. Default value : 1.

high_th - High Threshold value used in Hysteresis Thresholding. Ranges from 1 to 10.When Hysteresis threshold is not selected, used as a threshold value for ordinary thresholding.

low_th - Low Threshold value used in hysteresis thresholding. Ranges from 0 to 1.

Output Parameter Include :

out_img - Output boiecox edge detected image.

DESCRIPTION

The function performs the Boiecox edge detection on an image.The Boiecox edge detection algorithm was developed by R.A.Boie and I.J.Cox in 1987. The algorithm is similar to that of Canny's edge detection algorithm.It is a multistep algorithm where the image is first blurred with a Gaussian kernel and then a matched filter is applied to the image.The matched filter is used to find and detect the edges in an image and localize them to subpixel accuracy.The image is then operated by a wiener filter which boosts the signal amplitude and reduces the noise that may be present in the image.Matched filter is applied in an edge normal direction whereas wiener filter is applied along the edge contour.

REFERENCE

1.R.A. Boie, I.J. Cox,Two Dimensional Optimum Recognition Using Matched Filters and Wiener Filters for Machine Vision,IEEE First International Conference on Computer Vision, IEEE,pp.450-456, 1987.

EXAMPLE

% Read Image

 input_img1 = imread('butterfly.tif');
 input_img2 = imread('car.bmp');

% Gaussian variance range from 0.5 to 10.

 va1 = 1;
 va2 = 2;

% Hysterisis thresholding.

 hys_th1 = 1;
 hys_th2 = 0;

% Morphological thinning.

 thin1 = 1;
 thin2 = 0;

% High threshold used in hysterisis thresholding.

 high_th1 = 1;
 high_th2= 2;

% Low threshold used in hysterisis thresholding.

 low_th1 = 0.2;
 low_th2 = 0.4;

% Call function

 output1 = boiecox_ed_cvip(input_img1,va1,hys_th1,thin1,high_th1,low_th1);
 output2 = boiecox_ed_cvip(input_img2,va1,hys_th2,thin2,high_th2,low_th2);

% Display input image

 figure;imshow(input_img1);title('Input Image1');
 figure;imshow(input_img2);title('Input Image2');

% Dispaly Output image

  figure; imshow(output1);title('Boiecox edge detected Image');
  figure; imshow(output2);title('Boiecox edge detected Image');

BOIECOX_ED_CVIP

Contents

SYNTAX

DESCRIPTION

REFERENCE

EXAMPLE

CREDITS