Skip to content

Instantly share code, notes, and snippets.

@fiskurgit
Created December 11, 2019 17:32
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save fiskurgit/e6f55848309494b1a8fbbd6e2913a302 to your computer and use it in GitHub Desktop.
Save fiskurgit/e6f55848309494b1a8fbbd6e2913a302 to your computer and use it in GitHub Desktop.
from PIL import Image
from pathlib import Path
import random
def gray_value(r, g, b):
return 0.2989 * r + 0.5870 * g + 0.1140 * b
def filter_2x2_bayer(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
matrix = [[1, 3], [4, 2]]
for x in range(width):
for y in range(height):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
gray += gray * matrix[x % 2][y % 2] / 5
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_3x3_bayer(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
matrix = [[3, 7, 4], [6, 1, 9], [2, 8, 5]]
for x in range(width):
for y in range(height):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
gray += gray * matrix[x % 3][y % 3] / 10
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_4x4_bayer(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
matrix = [[1, 9, 3, 11], [13, 5, 15, 7], [4, 12, 2, 10], [16, 8, 14, 6]]
for x in range(width):
for y in range(height):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
gray += gray * matrix[x % 4][y % 4] / 17
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_5x3_bayer(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
matrix = [[9, 3, 0, 6, 12], [10, 4, 1, 7, 13], [11, 5, 2, 8, 14]]
for x in range(width):
for y in range(height):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
# Horizontal:
gray += gray * matrix[x % 3][y % 5] / 16
# Vertical:
# gray += gray * matrix[y % 3][x % 5] / 16
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_8x8_bayer(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
matrix = [[24, 10, 12, 26, 35, 47, 49, 37],
[8, 0, 2, 14, 45, 59, 61, 51],
[22, 6, 4, 16, 43, 57, 63, 53],
[30, 20, 18, 28, 33, 41, 55, 39],
[34, 46, 48, 36, 25, 11, 13, 27],
[44, 58, 60, 50, 9, 1, 3, 15],
[42, 56, 62, 52, 23, 7, 5, 17],
[32, 40, 54, 38, 31, 21, 19, 29]]
for x in range(width):
for y in range(height):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
gray += gray * matrix[x % 8][y % 8] / 65
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_floyd_steinberg(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
errors = [[0 for x in range(width)] for y in range(height)]
for x in range(width - 1):
for y in range(height - 1):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + errors[x][y] < threshold:
error = gray + errors[x][y]
filtered_image[x, y] = 0
else:
error = gray + errors[x][y] - 255
filtered_image[x, y] = (255, 255, 255)
errors[x + 1][y] += 7 * error / 16
errors[x - 1][y + 1] += 3 * error / 16
errors[x][y + 1] += 5 * error / 16
errors[x + 1][y + 1] += 1 * error / 16
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_jarvis_judice_ninke(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
errors = [[0 for x in range(width)] for y in range(height)]
for x in range(width - 1):
for y in range(height - 1):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + errors[x][y] < threshold:
error = gray + errors[x][y]
filtered_image[x, y] = 0
else:
error = gray + errors[x][y] - 255
filtered_image[x, y] = (255, 255, 255)
errors[x + 1][y] += 7 * error / 48
errors[x + 2][y] += 5 * error / 48
errors[x - 2][y + 1] += 3 * error / 48
errors[x - 1][y + 1] += 5 * error / 48
errors[x][y + 1] += 7 * error / 48
errors[x + 1][y + 1] += 5 * error / 48
errors[x + 2][y + 1] += 3 * error / 48
errors[x - 2][y + 2] += 1 * error / 48
errors[x - 1][y + 2] += 3 * error / 48
errors[x][y + 2] += 5 * error / 48
errors[x + 1][y + 2] += 3 * error / 48
errors[x + 2][y + 2] += 1 * error / 48
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_sierra(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
errors = [[0 for y in range(height)] for x in range(width)]
for x in range(width - 2):
for y in range(height - 2):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + errors[x][y] < threshold:
error = gray + errors[x][y]
filtered_image[x, y] = 0
else:
error = gray + errors[x][y] - 255
filtered_image[x, y] = (255, 255, 255)
errors[x + 1][y] += 5 * error / 32
errors[x + 2][y] += 3 * error / 32
errors[x - 2][y + 1] += 2 * error / 32
errors[x - 1][y + 1] += 4 * error / 32
errors[x][y + 1] += 5 * error / 32
errors[x + 1][y + 1] += 4 * error / 32
errors[x + 2][y + 1] += 2 * error / 32
errors[x - 1][y + 2] += 2 * error / 32
errors[x][y + 2] += 3 * error / 32
errors[x + 1][y + 2] += 2 * error / 32
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_two_row_sierra(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
errors = [[0 for y in range(height)] for x in range(width)]
for x in range(width - 2):
for y in range(height - 1):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + errors[x][y] < threshold:
error = gray + errors[x][y]
filtered_image[x, y] = 0
else:
error = gray + errors[x][y] - 255
filtered_image[x, y] = (255, 255, 255)
errors[x + 1][y] += 4 * error / 16
errors[x + 2][y] += 3 * error / 16
errors[x - 2][y + 1] += 1 * error / 16
errors[x - 1][y + 1] += 2 * error / 16
errors[x][y + 1] += 3 * error / 16
errors[x + 1][y + 1] += 2 * error / 16
errors[x + 2][y + 1] += 1 * error / 16
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_stucki(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
errors = [[0 for y in range(height)] for x in range(width)]
for x in range(width - 2):
for y in range(height - 2):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + errors[x][y] < threshold:
error = gray + errors[x][y]
filtered_image[x, y] = 0
else:
error = gray + errors[x][y] - 255
filtered_image[x, y] = (255, 255, 255)
errors[x + 1][y] += 8 * error / 42
errors[x + 2][y] += 4 * error / 42
errors[x - 2][y + 1] += 2 * error / 42
errors[x - 1][y + 1] += 4 * error / 42
errors[x][y + 1] += 8 * error / 42
errors[x + 1][y + 1] += 4 * error / 42
errors[x + 2][y + 1] += 2 * error / 42
errors[x - 2][y + 2] += 1 * error / 42
errors[x - 1][y + 2] += 2 * error / 42
errors[x][y + 2] += 4 * error / 42
errors[x + 1][y + 2] += 2 * error / 42
errors[x + 2][y + 2] += 1 * error / 42
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_atkinson(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
errors = [[0 for y in range(height)] for x in range(width)]
for x in range(width - 2):
for y in range(height - 2):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + errors[x][y] < threshold:
error = gray + errors[x][y]
filtered_image[x, y] = 0
else:
error = gray + errors[x][y] - 255
filtered_image[x, y] = (255, 255, 255)
errors[x + 1][y] += error / 8
errors[x + 2][y] += error / 8
errors[x - 1][y + 1] += error / 8
errors[x][y + 1] += error / 8
errors[x + 1][y + 1] += error / 8
errors[x][y + 2] += error / 8
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_left_to_right_error_diffusion(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
for y in range(height):
error = 0
for x in range(width):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray + error < threshold:
delta = gray
filtered_image[x, y] = 0
else:
delta = gray - 255
filtered_image[x, y] = (255, 255, 255)
if abs(delta) < 10:
delta = 0
error += delta
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_random(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
for y in range(height):
for x in range(width):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
threshold = (random.random() * 1000) % 256
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
def filter_threshold(image_file, threshold):
source_image = Image.open(image_file).convert("RGB")
image_copy = source_image.copy()
filtered_image = image_copy.load()
width, height = source_image.size
for y in range(height):
for x in range(width):
r, g, b = source_image.getpixel((x, y))
gray = gray_value(r, g, b)
if gray < threshold:
filtered_image[x, y] = 0
else:
filtered_image[x, y] = (255, 255, 255)
output_filename = Path(image_file.parent, 'processed_%s' % image_file.name)
image_copy.save(output_filename)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment