ShadowCast

class augraphy.augmentations.shadowcast.ShadowCast(shadow_side='random', shadow_vertices_range=(1, 20), shadow_width_range=(0.3, 0.8), shadow_height_range=(0.3, 0.8), shadow_color=(0, 0, 0), shadow_opacity_range=(0.2, 0.9), shadow_iterations_range=(1, 2), shadow_blur_kernel_range=(101, 301), p=1)

Bases: Augmentation

Emulates shadow effect on the surface of the paper.

Parameters:

• shadow_side (string, optional) – Side of the image to apply the shadow effect. Choose from “random”, “left”, “right”, “top” or “bottom”.

• shadow_vertices_range (Tuple, optional) – Pair of ints determining the vertices of shadow shape. The minumum value is 1, excluding the start point and end point to form a triangle.

• shadow_width_range (tuple, optional) – Pair of values determining the width of shadow effect. The width value will be in percentage of the image size if the value is float and in between 0.0 - 1.0: shadow_width (int) = image width * shadow_width (float and 0.0 - 1.0)

• shadow_height_range (tuple, optional) – Pair of values determining the maximum height of shadow effect. The height value will be in percentage of the image size if the value is float and in between 0.0 - 1.0: shadow_height (int) = image height * shadow_height (float and 0.0 - 1.0)

• shadow_color (tuple, optional) – Color of shadow in BGR or use “random” for random color.

• shadow_opacity_range (tuple, optional) – Pair of floats determining the opacity of the shadow. The minimum of opacity is 0.0 and the maximum value of opacity is 1.0.

• shadow_iterations_range (tuple, optional) – Pair of ints determining the iteration numbers to apply the shadow effect.

• shadow_blur_kernel_range (tuple, optional) – Pair of ints determining the value of the kernel to blur the shadow effect.

• p (float, optional) – The probability this Augmentation will be applied.

Overview

The ShadowCast augmentation applies random shadow effect into the image.

Initially, a clean image with single line of text is created.

Code example:

# import libraries
import cv2
import numpy as np
from augraphy import *


# create a clean image with single line of text
image = np.full((500, 1500,3), 250, dtype="uint8")
cv2.putText(
    image,
    "Lorem ipsum dolor sit amet, consectetur adipiscing elit",
    (80, 250),
    cv2.FONT_HERSHEY_SIMPLEX,
    1.5,
    0,
    3,
)

cv2.imshow("Input image", image)

Clean image:

Example 1

In this example, a ShadowCast augmentation instance is initialized and the shadow side is set to “bottom”. The shadow_vertices_range is set to any value between 2 and 3 (2,3) so that the final shadow shape is either a rectangle or pentagon. The width and height of the shadow is set to any value between 50% to 80% (0.5, 0.8) of the image size. The shadow color is set to black (0,0,0) and the opacity is set to any value between 50% and 60% (0.5, 0.6). The number of stacked shadow effect is set to random value in between 1 and 2 (1,2). The value of kernel to blur the shadow is set to a high value and the value is between 101 and 301 (101, 301).

Code example:

shadowcast = ShadowCast(shadow_side = "bottom",
                        shadow_vertices_range = (2, 3),
                        shadow_width_range=(0.5, 0.8),
                        shadow_height_range=(0.5, 0.8),
                        shadow_color = (0, 0, 0),
                        shadow_opacity_range=(0.5,0.6),
                        shadow_iterations_range = (1,2),
                        shadow_blur_kernel_range = (101, 301),
                        )

img_shadowcast = shadowcast(image)
cv2.imshow("shadowcast", img_shadowcast)

Augmented image: