pyimkernel 0.7.1

Applying some image kernel(s) on a grayscale or color scale image

Pyimkernel

With this package, You can apply various image kernels such as Blur, Sobel, Scharr and so forth (The list of image kernels is mentioned below) on a grayscale or color-scale image, and then show them. All of these happens using the "ApplyKernels" class to reach a wide range of effects and enhancements in digital images.

Installation

pip install pyimkernel

Usage

from pyimkernel import ApplyKernels
import mnist # pip install mnist
import cv2
import os

# Load data
X_train, X_test, y_train, y_test = mnist.train_images(), mnist.test_images(), mnist.train_labels(), mnist.test_labels()

# Create an instance
imkernel = ApplyKernels(random_seed=0)

# Grayscale
# Show image 9 
imkernel.imshow(X_train[19], cmap=plt.cm.gray)

# Apply blur kernel on a grayscale image 9
filtered_image = imkernel.apply_filter_on_gray_img(X_train[19], kernel_name='blur')

# Show the filtered image 9
imkernel.imshow(image=filtered_image, cmap='gray')

# the Color-Scale image
# Read the flower image
image1 = cv2.imread(os.path.join('Images', '1.jpg'))

# Show the flower image
imkernel.imshow(image1.reshape(image1.shape[0], -1), cmap='gray', figsize=(20, 10))

# Show the filtered flower image
blurred_image = imkernel.apply_filter_on_color_img(image1, kernel_name='motion blur', with_resize=True)
imkernel.imshow(image=imkernel.apply_filter_on_gray_img(blurred_image, kernel_name='sharpen'),
                figsize=(7, 6), cmap=plt.cm.gray)

The Grayscale Output

Before Applying the blur kernel on a grayscale image 9

After Applying the blur kernel on a grayscale image 9

The Color Scale Output

Before Applying the laplacian kernel on a color-scale flower image

After Applying the Sharpen Kernel on a filtered color-scale image using the Motion Blur Kernel and assigning True to the with_resize parameter

Image kernels

Image kernels are listed below:

• blur
• bottom sobel
• emboss
• identity
• left sobel
• outline
• right sobel
• sharpen
• top sobel
• horizontal edge
• vertical edge
• box blur
• laplacian
• prewitt horizontal edge
• prewitt vertical edge
• high-pass filter
• unsharp masking
• dilate
• soften
• scharr horizontal edge
• scharr vertical edge
• motion blur

All Tests Passed.