Applying pencil sketch transformation

With the tricks that we learned from the previous sections in our bag, we are now ready to take a look at the entire procedure.

The following procedure shows you how to convert a color image into grayscale. After that, we aim to blend the grayscale image with its blurred negative:

1. First, we convert an RGB image (imgRGB) into grayscale:

img_gray = cv2.cvtColor(img_rgb, cv2.COLOR_RGB2GRAY) 

As you can see, we have used cv2.COLOR_RGB2GRAY as a parameter to the cv2.cvtColor function, which changes the color spaces. Note that it does not matter whether the input image is RGB or BGR (which is the default for OpenCV); we will get a nice grayscale image in the end.

2. Then, we invert the image and blur it with a large Gaussian kernel of size (21,21):

    inv_gray = 255 - gray_image
    blurred_image = cv2.GaussianBlur(inv_gray, (21, 21), 0, 0)

3. We use dodge to blend the original grayscale image with the blurred inverse:

    gray_sketch = cv2.divide(gray_image, 255 - blurred_image, 
    scale=256)

The resulting image looks like this:

Did you notice that our code can be optimized further? Let's take a look at how to optimize with OpenCV next.