Canny Edge Detection

This process is the Digital image processing in OpenCV libary to find an edge image by using Canny algorithm.

using function.

void cvCanny( const CvArr* image, CvArr* edges, double threshold1,
double threshold2, int aperture_size=3 );

Example for use OpenCV libary.

IplImage *pCanny=cvCreateImage(cvGetSize(pimage),IPL_DEPTH_8U,1);
cvCanny (pGray,pCanny, 200,250, 3 );
cvNamedWindow("Canny",CV_WINDOW_AUTOSIZE);
cvShowImage("Canny", pCanny);

result.

Sobel edge detection on OpenCV

In this system, Finding the edge image by Sobel Edge Detection in Digital image process by OpenCV libary. That is find edge by use template size 3x3 with two templates to use the first template for the difference in the horizontal (Xdiff) and The difference in the vertical (Ydiff) by at

Y_difference(x,y) = value(x,y) - value(x,y+1)
X_difference(x,y) = value(x,y) - value(x-1,y)

example use function in OpenCV

//Convert to Sobel
IplImage *pSobel=cvCreateImage(cvGetSize(pimage),IPL_DEPTH_16S,1);
cvSobel( pGray,pSobel,1,1,7 );
cvNamedWindow("Sobel",CV_WINDOW_AUTOSIZE);
cvShowImage("Sobel", pSobel);

result:

Template for Edge detection

To find the edge image horizontally simple in Image processing. How it is to find the difference between one point on the address below (or above) the following point.

Ydiff(x,y) = I(x,,y) - I(x,y+1)------------ (1)

The Ydiff is the difference in axial location of image and I (x,, y) is the intensity of the pixel position (x,, y) in image results of using equation (1) is equivalent to. image convolution with template.
1
-1

Finding the vertical edge image can be the same.

Xdiff(x,,y) = I(x,,y) - I(x-,y) -------------(2)

The Xdiff is the difference in horizontal and equation (2) is equivalent to convolution with the template of image processing.

Edge Detection

Find the perimeter of objects in the picture. With that in mind a line around the object. We can calculate area (size) or know the type of objects that need. However, finding the right edge of picture is not complete as easy. In particular, finding the edge of the picture quality is less difference between foreground and background, low Or brightness is not always around the image edge image resulting from the difference in intensity from one point to another point if it is different this valuable edge image will be visible. If the difference is less. Edge images is not clear.