Find Image Gradients Using Scharr Operator in OpenCV Python

Using the Scharr operator, we can compute image gradients in horizontal as well as vertical direction using first order derivatives. The gradients are computed for a grayscale image. You can apply Scharr operation on an image using the method cv2.scharr().

Syntax

The following syntax is used to compute the image gradients using Scharr derivative ?

cv2.Scharr(img, ddepth, xorder, yorder)

Parameters

• img ? The original input image

• ddepth ? Desired depth of the output image. It has information about what kind of data is stored in the output image. We use cv2.CV_64F to as ddepth. It is a 64bit floating-point opencv

• xorder ? The order of derivatives in horizontal direction (X-direction). Set xorder=1, yorder=0 for the 1st order derivative in X-direction.

• Yorder ? The order of derivatives in vertical direction (Y-direction). Set xorder=0, yorder=1 for the 1st order derivative in Y-direction.

Steps

You can use the following steps to find image gradients using Scharr derivative ?

Import the required library. In all the following Python examples, the required Python library is OpenCV. Make sure you have already installed it.

import cv2

Read the input image using cv2.imread() as a grayscale image.

img = cv2.imread('lines.jpg',0)

Compute the Sobel or Laplacian derivative using cv2.Scharr(). This derivative refers to the image gradient.

scharrx = cv2.Scharr(img,cv2.CV_64F,1,0)

Display the image gradient using cv2.imshow() method.

cv2.imshow("Scharr X", scharrx)
cv2.waitKey(0)
cv2.destroyAllWindows()

Let's have a look at some examples for more clear understanding.

Example 1

In the python example below, we compute the image gradient using the Scharr operator in X (horizontal) as well as Y (vertical) directions.

# import required libraries
import cv2

# read the input image as a grayscale image
img = cv2.imread('window.jpg',0)

# compute the 1st order Sobel derivative in x direction
scharrx = cv2.Scharr(img,cv2.CV_64F,1,0)

# compute the 1st order Sobel derivative in y direction
scharry = cv2.Scharr(img,cv2.CV_64F,0,1)

# display scharrx and scharry
cv2.imshow("Scharr X", scharrx)
cv2.waitKey(0)
cv2.imshow("Scharr Y", scharry)
cv2.waitKey(0)
cv2.destroyAllWindows()

We will use the following image "window.jpg" as the Input File in the above program.

Output

When you execute the above program, it will produce the following two output windows? "Scharr X" and "Scharr Y".

Example 2

In the Python example below, we compute the image gradient using Scharr operator in X (horizontal) as well as Y (vertical) directions.

# import required libraries
import cv2

# read the input image as a grayscale image
img = cv2.imread('tutorialspoint.png',0)

# compute the 1st order Sobel derivative in x direction
scharrx = cv2.Scharr(img,cv2.CV_64F,1,0)

# compute the 1st order Sobel derivative in y direction
scharry = cv2.Scharr(img,cv2.CV_64F,0,1)

# display scharrx and scharry
cv2.imshow("Scharr X", scharrx)
cv2.waitKey(0)
cv2.imshow("Scharr Y", scharry)
cv2.waitKey(0)
cv2.destroyAllWindows()

We will use this image "tutorialspoint.png" as the Input File in the above program.

Output

The above program on execution will produce the following two output windows