Name: Catherine Gai

SID: 3034712396

Email: catherine_gai@berkeley.edu

Link to project report website: https://inst.eecs.berkeley.edu/~cs194-26/fa21/upload/files/proj2/cs194-26-aay/catherine_gai_proj2/Project%2002.html

This folder contains four functional python files: "filter.py", "frequency.py", "align_image_code.py", "hybrid_image_starter.py", "stack.py", "blending.py".

The folder also contains extra image files: "big_sur.jpeg","rainbow.jpeg" (for image sharpening); "angjoo.jpeg", "efros.jpeg", "wolf.jpeg", "lion.jpeg" (for image hybrid); "pizza1.jpeg", "pizza2.jpeg", "left_squirrel.jpeg", "right_squirrel.jpeg" (for image blending).

filter.py:

This python file contains functions and other python commands adequate to generate all required images for Part 1.

• derivative(params): calculates the x-derivative and y-derivative of a given image.

To stimulate image-generation process, run python filter.py. This command will result in seven images. Sequentially they are: x-derivative of the original image, y-derivative of the original image, blurred image, x-derivative of the blurred image by applying Gaussian first then the derivative, y-derivative of the blurred image by applying Gaussian first then the derivative, x-derivative of the blurred image by directly applying derivative of Gaussian filter, y-derivative of the blurred image by directly applying derivative of Gaussian filter.

frequency.py:

This python file contains functions for image sharpening.

• sharpen(params): sharpens the given image with respect to specified Gaussian filter and weight.
• main(): sharpens "taj.jpeg", "big_sur.jpeg". First blurs then sharpens "rainbow.jpeg".

To stimulate image-generation process, run python frequency.py. This command will generate five images. Sequentially they are: sharpened "taj.jpeg", sharpened "big_sur.jpeg", blurred "rainbow.jpeg", sharpened "rainbow.jpeg" after blurring.

align_image_code.py:

The code is provided by the staff to facilitate image hybrid.

hybrid_image_starter.py:

This python file contains functions that perform image hybrid.

• hybrid_image_bw(params): performs image hybrid on two gray-scale images, Image 1 is turned to low-frequency version to be visible only from faraway. Image 2 si turned to high-frequency verison to be visible only from near. The function also shows frequency distribution of filtered low-frequency and high-frequency images.
• Hybrid_image_color(params): performs image hybrid on two colored images, Image 1 is turned to low-frequency version to be visible only from faraway. Image 2 si turned to high-frequency verison to be visible only from near.
• hybrid_cat(): performs image hybrid on "DerekPicture.jpeg" and "nutmeg.jpeg". The function outputs both gray-scale results and colored results. It also shows the frequency distribution of the hybrid image. The hybrid image shows a cat while viewing from near and Derek while viewing from faraway.
• hybrid_efros(): performs image hybrid on "efros.jpeg" and "kanazawa.jpeg". The function outputs both gray-scale results. It also shows the frequency distribution of the hybrid image. The hybrid image shows Kanazawa while viewing from near and Efros while viewing from faraway.
• hybrid_wolf(): performs image hybrid on "lion.jpeg" and "wolf.jpeg". The function outputs both gray-scale results. It also shows frequency distribution of the two original images, those of the filtered images, and that of the hybrid image. The hybrid image shows a lion while viewing from near and a wolf while viewing from faraway.

To stimulate image-generation process, run python hybrid_image_starter.py.

stack.py:

This function contains functions that generates Gaussian stack and Laplacian stack.

• gaussian_stack(params): generates Gaussian stack of an image, given the number of layers of that Gaussian stack and a factor that determines the size of Gaussian kernels at each level.
• laplacian_stack(params): generates Laplacian stack of an image, given the number of layers of that Laplacian stack and a factor that determines the size of Gaussian kernels at each level. Calls gaussian_stack(params), and use its returned results to calculate corresponding Laplacian stack.
• main(): Calculates the 6-layer Gaussian stack and 6-layer Laplacian stack of "oraple.jpeg".

To stimulate image-generation process, run python stack.py.

blending.py:

This function contains functions that blends two images together.

• blending(params): given two images, a blending mask, laplacian_factor that is used to create Laplacian stack for those input images, gaussian_factor that is used to create Gaussian stack for the mask, output the blended result.
• blending_apple_color(): blends colored "apple.jpeg" and "orange.jpeg" together.
• blending_apple_gray(): blends gray_scale "apple.jpeg" and "orange.jpeg" together.
• blending_pizza(): blends colored "pizza1.jpeg" and "pizza2.jpeg" together.
• blending_squirrels(): blends colored "left_squirrel.jpeg" and "right_squirrel.jpeg" together. This function uses irregular mask that is a combination of a square and a circle.

To stimulate image-generation process, run python blending.py.