Zeitraffer - ein Foto pro Tag

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81

import imageio as iio
import numpy as np
import os
import sys
import skimage.filters
import skimage.exposure

def build_filter(maskname, maxlen):

    mask = iio.imread(maskname).astype(float) / 255.0 # values 0..1 expected
    circles = []
    for i in range(0,maxlen):
        # The timetunnel mask files center_0.jpg ... center_8.jpg must be present in the workdir
        image = iio.imread(f'center_{i}.jpg').astype(float)
        image = skimage.filters.gaussian(image, sigma=(30,30), multichannel=True)
        image = image*(float(maxlen-i)/255.0)
        circles.append(image)
        # write blurred just for debugging / demonstration
        #iio.imwrite(f'center_blurred_{i}.jpg',convert(image, 0, 255, np.uint8))

    return mask, circles

def convert(img, target_type_min, target_type_max, target_type):
    imin = img.min()
    imax = img.max()

    a = (target_type_max - target_type_min) / (imax - imin)
    b = target_type_max - a * imax
    new_img = (a * img + b).astype(target_type)
    return new_img


def main():
    MAXLEN = 9
    MAXSTEPS = 3
    outdir = 'frames'

    # initialize filters and history
    mask_off, circles = build_filter(sys.argv[1], MAXLEN)
    mask_on = 1-mask_off # values in range 0..1
    mask_off = mask_off + 1
    history = np.zeros((MAXLEN,)+mask_off.shape)

    for idx, path in enumerate(sys.argv[2:]):
        today = iio.imread(path).astype(float)


        # drop oldest image, insert newest at pos 0
        # summed_images : [ today, today-1 (yesterday), today -2, today -3 , ...]
        history = np.roll(history,1,0)
        history[0]=today
        yesterday = history[1]

        # et filename without path and ext, and ext separately
        name, ext = os.path.splitext(os.path.normpath(path).split(os.sep)[-1])

        print(f'{name}{ext}')

        if idx >= MAXLEN:
            # at this point, history is filled.
            # let's build frames

            # the interpolated image should contain a weighted average of the history images
            # instead of a scalar weight, we use masks containing weighted blurred circles
            # this would give us a subtle timetunnel effect

            interpolated = np.zeros(today.shape)
            for i in range(0,MAXLEN):
                interpolated += history[i] * circles[i]

            # circle_0 values range from 0..9, circle_8  from 0..1.
            # The mean is approx MAXLEN**2/2.0.
            # we must normalize the values to 0..1
            interpolated /= (MAXLEN**2/2.0)

            for step in range(0,MAXSTEPS):
                # for each day we build MAXSTEPS frames
                # we add to the interpolated image a masked fading between the images of yesterday and today
                # This gives a nice emphasis on the area specified ba mask
                frame = mask_off*interpolated + mask_on*( (step/MAXSTEPS ) * today + ((MAXSTEPS-step)/MAXSTEPS)*yesterday)
                iio.imwrite(f'{outdir}/{name}_{step}{ext}',convert(frame, 0, 255, np.uint8))

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

# create video
rm frames.mp4 || true
ffmpeg -f image2 -r 30 -pattern_type glob -i 'frames/*.jpg' -vcodec libx264 -acodec aac -vf "pad=ceil(iw/2)*2:ceil(ih/2)*2"  frames.mp4

# add audio
rm audio.mp4 || true
ffmpeg -i frames.mp4 -i slide.mp3 -map 0 -map 1:a -c:v copy -shortest audio.mp4

# fade
dur=$(ffprobe -loglevel error -show_entries format=duration -of default=nk=1:nw=1 "audio.mp4")
voff=$(bc -l <<< "$dur"-3)
aoff=$(bc -l <<< "$dur"-7)
rm video.mp4 || true
ffmpeg  -i audio.mp4 -vf "fade=t=in:st=0:d=3, fade=t=out:st=$voff:d=3" -af "afade=t=out:st=$aoff:d=7" video.mp4