Last active
October 19, 2018 23:06
-
-
Save badjano/bcc2b13e99707542d2708c97b5759148 to your computer and use it in GitHub Desktop.
ImageConvolutionImage
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from PIL import ImageFilter, Image | |
from math import floor, sqrt, tanh | |
from threading import Thread, Lock | |
from multiprocessing import cpu_count | |
from queue import Queue | |
import time | |
import pygame | |
lock = Lock() | |
q = Queue() | |
threads_status = {} | |
def clamp(a, _min, _max): | |
return max(_min, min(a, _max - 1)) | |
def limit(point, w, h): | |
return ( | |
clamp(point[0], 0, w), | |
clamp(point[1], 0, h) | |
) | |
def conv_image(im, kernel, out, rect, quad_index, t): | |
im_w, im_h = im.size | |
k_w, k_h = kernel.size | |
x_start, y_start, x_end, y_end = rect | |
i = 0 | |
total_pixels = (x_end - x_start) * (y_end - y_start) | |
for x in range(x_start, x_end): | |
for y in range(y_start, y_end): | |
c = (0, 0, 0) | |
total = (0, 0, 0) | |
for x2 in range(k_w): | |
for y2 in range(k_h): | |
a = floor(x2 / 2) | |
b = floor(y2 / 2) | |
p = limit((x - a, y - b), im_w, im_h) | |
r, g, b = im.getpixel(p) | |
r2, g2, b2 = kernel.getpixel((x2, y2)) | |
r2 = r2 * 2 - 128 | |
g2 = g2 * 2 - 128 | |
b2 = b2 * 2 - 128 | |
c = (c[0] + r * r2, c[1] + g * g2, c[2] + b * b2) | |
total = (total[0] + abs(r2), total[1] + abs(g2), total[2] + abs(b2)) | |
pixel = ( | |
int(c[0] / total[0]), | |
int(c[1] / total[1]), | |
int(c[2] / total[2]) | |
) | |
out.putpixel((x, y), pixel) | |
i += 1 | |
threads_status[quad_index] = i / total_pixels | |
def do_work(data, t): | |
im, kernel, out, rect, quad_index = data | |
conv_image(im, kernel, out, rect, quad_index, t) | |
# lock.acquire() | |
# print("Done on thread: ",t) | |
# lock.release() | |
def worker(data): | |
thread_index = q.get() | |
while len(data) > 0: | |
do_work(data.pop(0), thread_index) | |
q.task_done() | |
def tiles(size, count): | |
ts = [] | |
side = round(sqrt(count)) | |
w, h = (round(size[0] / side), round(size[1] / side)) | |
for i in range(count): | |
x = (i % side) * w | |
y = floor(i / side) * h | |
ts.append((x, y, x + w, y + h)) | |
return ts | |
def gcd(a, b): | |
if b == 0: | |
return a | |
else: | |
return gcd(b, a % b) | |
def monitor(): | |
done = False | |
while not done or (len(data) > 0 and sum([v for k, v in threads_status.items()]) < len(threads_status.items())): | |
print(len(data)) | |
print(", ".join(["%d : %.02f" % (k, v) for k, v in threads_status.items()])) | |
data = [] | |
def conv_iteration(): | |
size_scale = 5 | |
size = (160 * size_scale, 90 * size_scale) | |
screen = pygame.display.set_mode(size) | |
pygame.display.set_caption("My Game") | |
done = False | |
clock = pygame.time.Clock() | |
im = Image.open("img/image.jpg").resize(size) | |
kernel = Image.open("img/kernel_abstract.png").convert("RGB") | |
out = Image.new('RGB', size, 0xffffff) | |
quads = gcd(size[0], size[1]) ** 2 | |
print(quads) | |
quad_index = 0 | |
all_tiles = tiles(size, quads) | |
for tile in all_tiles: | |
data.append((im, kernel, out, tile, quad_index)) | |
threads_status[quad_index] = 0 | |
quad_index += 1 | |
threads = [] | |
for i in range(cpu_count() - 2): | |
q.put(i) | |
t = Thread(target=worker, args=(data,)) | |
t.daemon = True | |
t.start() | |
threads.append(t) | |
while not done and sum([1 for t in threads if t.isAlive()]) > 0: | |
for event in pygame.event.get(): | |
if event.type == pygame.QUIT: | |
done = True | |
screen.fill((0, 0, 0)) | |
for k, v in threads_status.items(): | |
tile = all_tiles[k] | |
pygame.draw.rect(screen, (127, 127, 127), [tile[0], tile[1], tile[2] - tile[0], 10]) | |
pygame.draw.rect(screen, (255, 255, 255), [tile[0], tile[1], v * (tile[2] - tile[0]), 10]) | |
pygame.display.flip() | |
clock.tick(60) | |
pygame.quit() | |
out.save("threaded_abstract.png") | |
conv_iteration() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment