DGNum/Stirling-PDF
15
0
Fork 0
Code Issues 4 Pull requests 1 Projects Releases Packages Wiki Activity

Merge pull request #588 from Emad-Eldin-G/patch-1

Browse source 
Optimized the code in detect-blank-pages.py
This commit is contained in:
Anthony Stirling 2023-12-28 07:55:25 +00:00 committed by GitHub
commit e4fb64ce16
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 8 additions and 18 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

20
scripts/detect-blank-pages.py
View file

@ -1,6 +1,7 @@
import cv2
import sys
import argparse
import numpy as np
def is_blank_image(image_path, threshold=10, white_percent=99, white_value=255, blur_size=5):
image = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
@ -15,19 +16,11 @@ def is_blank_image(image_path, threshold=10, white_percent=99, white_value=255,
_, thresholded_image = cv2.threshold(blurred_image, white_value - threshold, white_value, cv2.THRESH_BINARY)
# Calculate the percentage of white pixels in the thresholded image
white_pixels = 0
total_pixels = thresholded_image.size
for i in range(0, thresholded_image.shape[0], 2):
for j in range(0, thresholded_image.shape[1], 2):
if thresholded_image[i, j] == white_value:
white_pixels += 1
white_pixel_percentage = (white_pixels / (i * thresholded_image.shape[1] + j + 1)) * 100
if white_pixel_percentage < white_percent:
return False
white_pixels = np.sum(thresholded_image == white_value)
white_pixel_percentage = (white_pixels / thresholded_image.size) * 100
print(f"Page has white pixel percent of {white_pixel_percentage}")
return True
return white_pixel_percentage >= white_percent
if __name__ == "__main__":
@ -39,9 +32,6 @@ if __name__ == "__main__":
blank = is_blank_image(args.image_path, args.threshold, args.white_percent)
if blank:
# Return code 1: The image is considered blank.
sys.exit(1)
else:
# Return code 0: The image is not considered blank.
sys.exit(0)
sys.exit(int(blank))