import os
import cv2
import numpy as np
from scipy.stats import mode
import time
import concurrent.futures
'''
multi-process to crop pictures.
'''
def crop(file_path_list):
origin_path, save_path = file_path_list
img = cv2.imread(origin_path)
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
closed_1 = cv2.erode(gray, None, iterations=4)
closed_1 = cv2.dilate(closed_1, None, iterations=4)
blurred = cv2.blur(closed_1, (9, 9))
# get the most frequent pixel
num = mode(blurred.flat)[0][0] + 1
# the threshold depends on the mode of your images' pixels
num = num if num <= 30 else 1
_, thresh = cv2.threshold(blurred, num, 255, cv2.THRESH_BINARY)
# you can control the size of kernel according your need.
kernel = np.ones((13, 13), np.uint8)
closed_2 = cv2.erode(thresh, kernel, iterations=4)
closed_2 = cv2.dilate(closed_2, kernel, iterations=4)
_, cnts, _ = cv2.findContours(closed_2.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
c = sorted(cnts, key=cv2.contourArea, reverse=True)[0]
# compute the rotated bounding box of the largest contour
rect = cv2.minAreaRect(c)
box = np.int0(cv2.boxPoints(rect))
# draw a bounding box arounded the detected barcode and display the image
# cv2.drawContours(img, [box], -1, (0, 255, 0), 3)
# cv2.imshow("Image", img)
# cv2.imwrite("pic.jpg", img)
# cv2.waitKey(0)
xs = [i[0] for i in box]
ys = [i[1] for i in box]
x1 = min(xs)
x2 = max(xs)
y1 = min(ys)
y2 = max(ys)
height = y2 - y1
width = x2 - x1
crop_img = img[y1:y1 + height, x1:x1 + width]
cv2.imwrite(save_path, crop_img)
# cv2.imshow("Image", crop_img)
# cv2.waitKey(0)
print(f'the {origin_path} finish crop, most frequent pixel is {num}')
def multi_process_crop(input_dir):
with concurrent.futures.ProcessPoolExecutor() as executor:
executor.map(crop, input_dir)
if __name__ == "__main__":
data_dir = ''
save_dir = ''
path_list = [(os.path.join(data_dir, o), os.path.join(save_dir, o)) for o in os.listdir(data_dir)]
start = time.time()
multi_process_crop(path_list)
print(f'Total cost {time.time()-start} seconds')
