![PDF PNG
ImageMagick (
import subprocess
png_path = os.path.join(output_path, 'image-%03d.png')
argv = ['convert',
'-deskew', '40%', #
'-density', str(DENSITY), # 72dpi
#
'-geometry', '{0}x{0}'.format(MAX_HEIGHT),
pdf_path, png_path]
#
if subprocess.call(argv) != 0:
print('failed: {}'.format(pdf_path))
24](https://image.slidesharecdn.com/30-41-160923081804/85/Python-4-24-320.jpg)
![NumPy
width, height = 832, 1242
(322ms )
column_avg_list = []
for xi in range(width):
sum = 0
for yi in range(height):
sum += img[yi, xi]
column_avg_list.append(sum / height)
(1.02ms )
import numpy as np
#
column_avg_list = np.average(img, axis=0) 30](https://image.slidesharecdn.com/30-41-160923081804/85/Python-4-30-320.jpg)
Pythonで実現する4コマ漫画の分析・評論
- 1. Python 4 PyConJp 2016 09.22 @esuji repo. https://github.com/esuji5/yonkoma2data
- 2. 4 4 Python ✌ ('ω' ✌ ) ✌ ('ω') ✌ ( ✌ 'ω') ✌ 2
- 3. 4 Python 4 ( ) ( ) 3
- 4. S @esuji WEB subcatalog ( ) WEB 4
- 5. 4 5
- 6. 4 G 6
- 7. 4 1. 5M vol.4 (2010 ) 4 2. 3. 4 4. (300 ) 5. 7
- 8. 2008 2016 9 8 3 kindle unlimited 2013 4 6 24:30 (14 ) 8
- 10. 10
- 11. ( 4 ) 11
- 12. 12
- 13. 8 13
- 14. Python WEB 14
- 15. Python 15
- 16. Python 16
- 17. C++ 17
- 18. Python 18
- 19. 4 Python ) 19
- 20. Python 4 20
- 21. 4 21
- 22. ( ) PDF PNG PNG 1 22
- 23. Python3 OpenCV 3.0.0 ImageMagick Google Cloud Vision API 23
- 24. PDF PNG ImageMagick ( import subprocess png_path = os.path.join(output_path, 'image-%03d.png') argv = ['convert', '-deskew', '40%', # '-density', str(DENSITY), # 72dpi # '-geometry', '{0}x{0}'.format(MAX_HEIGHT), pdf_path, png_path] # if subprocess.call(argv) != 0: print('failed: {}'.format(pdf_path)) 24
- 25. PNG 1 25
- 27. 27
- 28. 28
- 29. 29
- 30. NumPy width, height = 832, 1242 (322ms ) column_avg_list = [] for xi in range(width): sum = 0 for yi in range(height): sum += img[yi, xi] column_avg_list.append(sum / height) (1.02ms ) import numpy as np # column_avg_list = np.average(img, axis=0) 30
- 31. Google Cloud Vision API API 31
- 32. 32
- 33. -1 … 33
- 34. -2 34
- 35. -3 35
- 36. 36
- 37. -4 37
- 38. -5 38
- 39. 39
- 40. CNN(Convolutional Neural Network) ( ) 40
- 41. 1 8 3 kindle unlimited BD-BOX 41