Canny Edge Detection and LoG difference

image-processing

Solution

The Canny edge detector is based on the first derivative of the image (Sobel mask convolution in x and y direction). The magnitude of the derivative will look like this:

You see that with this operation lines can be identified by pixels which have a high value (are white). The canny algorithm will then also apply non-maxima suppresion and line tracing (see Wikipedia).

The Laplacian of Gaussian operator however, is based on the second derivative of the image. A response of this operator will look like this:

The highest response of the LoG operator will be at the center of blob-like structures in images (same size as the LoG kernel). Lines can be identified using the LoG operator not by finding high magnitudes in the image but by inflection points (zero-crossings). You can see that each line is on one side dark and on the other white. Based on the second derivative, it is highly sensitive to noise.

Problem

What is the difference between Canny Edge Detector and LoG(Laplace of Gaussian) in Diagital Image Processing ?? I am totally new to Digital Image Processing .. I would be thankful if someone could tell me the difference . Thanks and Regards.

Original source