ASCII_Art_Generator

What is ASCII_ART_Generator?

ASCII_ART_Generator is a program that generates ASCII Arts from given images.

You might wonder, how it works?

Well, first you have to understand how images works, images are stored in comupters in 3 layered 2d grid/array, each layer representing RGB (Red, Green, Blue) color intensity

This is how images are stored in computer.

As images are stored in grid, their indices represent pixels in an image

Now lets try to understand how grayscale (black & white) images are stored in computer, these can be stored and understood easily, no complex 3 layered architectured.

The above image demonstrates how grayscale images are stored in computers, each pixel is an element stored in a 2D array, representing the brightness intensity of that pixel (from 0 to 225).

Technical information

Now you know how images are stored in computers, and this is what I used for creating ASCII Art.

Here images are of 2 types, colored which contain 3 layers i.e RGB and another is Grayscale i.e Black and white image

Grayscale image's index hold the curresponding brightness intensity of that pixel in UINT8 format ranging from 0 to 255, 0 being the lowest brightness and 255 being highest brightness.

As brightness and complete picture is available in indices (image[100][100]) and their brightness value ranging from 0 to 255, using numpy we can configure these parameteres leading to image manipulation, in my code I have also reduced the brightness of image with 50 for better contrasted result, these can be adjusted accordingly

image = cv.imread(input_path, cv.IMREAD_GRAYSCALE)
image = image.astype(np.int16) - 50
image = np.clip(image, 0, 255)
image = image.astype(np.uint8)

Here image is in UINT8 format, for manipulating values and to avoid overflow or underflow within range 0 and 255, we first convert it to INT16 increasing the range for manipulation and then clipping it for the desired range.

For creating ASCII Art, we can simply replace pixels with Ascii characters (Higher brightness intensity can be represented with dense and bigger AScii characters such as "@", "#", "$" etc and lower can be represented as ".", "'", "," etc

To represent the gradient of brightness within image, we would need more character

List of characters: "$@B%8&WM#*oahkbdpqwmZO0QLCJUYXzcvunxrjft/|()1{}[]?-_+~<>i!lI;:,"^`'. "

Now these characters are assigned to respective brightness intensity and then is printed on image and then saved

How program works

Reading image using OpenCV

image = cv.imread(input_path, cv.IMREAD_GRAYSCALE)

then adjusting brightness according to the taste.

image = image.astype(np.int16) - 50
image = np.clip(image, 0, 255)
image = image.astype(np.uint8)

Creating complete black image same as the size of provided image

(height, width) = image.shape
new_image = np.zeros((height, width), dtype = np.uint8)

Main printing of characters on new image

for y in range(0, height, char_h): #skipped by char_h as height of a character
        for x in range(0, width, char_w): #skipped by char_w as width of a character
            brightness = image[y, x] #brightness intensity of image at (x, y)(y - height index, x - width index)
            character = brightness2ascii(brightness) #brightness to ASCII character conversion
            #cv.putText function to write text (ASCII Characters) on image
            cv.putText(new_image, #Image to write on
                       character, #str to write
                       (x, y + char_h - 2), #Coordinate/position
                       font, #Font
                       font_scale, #Font scale
                       (255,255,255), #White color of character
                       thickness, #thickness of character
                       lineType=cv.LINE_AA) #parameter the determines how edges are rendered (LINE_AA = Auto Aliasing)

brightness2ascii() function

def brightness2ascii(brightness):
    character_index = int((brightness/255) * (len(ASCII_CHARS) - 1))
    return ASCII_CHARS[character_index]

It calculates the index of ascii character and returns the character

How to use it

First you would need python installed in your machine and dependencies (Numpy, OpenCV) installed

pip install numpy
pip install opencv-python

Instruction: Use higher resolution images (4k+) captured from high end devices. (Higher the resolution, better will be the quality)
Enter extension and image name correctly.
Using large size of image (size > 30Mb) might slow down the process as it takes time, so keep the patience.
Avoid photos with low contrast or higher brightness.

Get Going:
Clone this repo in your local machine, paste some images in the same directory as main.py, run main.py and enter the name of image with extension and boom, you got your own ASCII Art generated!!
You can also enter the path of image where it is stored for generating ASCII art, ASCII art will be generated in the current directory of main.py.

Sources:

OpenCV documentation - Python
https://docs.opencv.org/4.x/d6/d00/tutorial_py_root.html
Numpy Documentation
https://numpy.org/doc/

Thanks for reading through, update any information or mistakes, happy coding and happy learning!