Sobel Edge

Block Category: Detection

Input Image Format

Input Image Requirements: None

Description: The Sobel Edge block detects the edges of an image in both the horizontal and vertical directions with the detected edges bright on a dark background. The Sobel operator is a discrete differentiation operator. It combines Gaussian smoothing and differentiation, and computes an approximation of the gradient of the image intensity function.

Delta: Specifies a value used to vary the pixel intensity. The delta value is added to the results prior to storing them in the output.

Derivative: Indicates the direction of edges to be detected. There are three supported combinations:

Horizontal derivative

Vertical derivative




Horizontal edges (y-gradient) detected



Vertical edges (x-gradient) detected



Calculates an approximation of the gradient (x-gradient and y-gradient) by combing both the horizontal and vertical derivatives

Horizontal: Indicates an integer variable representing the x-derivative.

Vertical: Indicates an integer variable reprenting the y-derivative.

Kernel Size: Specifies a value used to extend the Sobel kernel. The value must be odd and positive and less than 32.

Scale: Indicates the scale value for the computed derivative values. The scale value helps to visualize a derivative in an image.


In this example, three different outputs are obtained by changing the derivative value.