Politrees commited on
Commit
1c8af3d
·
verified ·
1 Parent(s): 54e567d

Update txt2spec.py

Browse files
Files changed (1) hide show
  1. txt2spec.py +29 -29
txt2spec.py CHANGED
@@ -7,35 +7,35 @@ import gradio as gr
7
  import soundfile as sf
8
  import os
9
 
10
- # Функция для создания изображения спектрограммы с текстом
11
  def text_to_spectrogram_image(text, base_width=512, height=256, max_font_size=80, margin=10, letter_spacing=5):
12
- # Шрифт и размер текста
13
  font_path = "/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf"
14
  if os.path.exists(font_path):
15
  font = ImageFont.truetype(font_path, max_font_size)
16
  else:
17
  font = ImageFont.load_default()
18
 
19
- # Определяем ширину текста с учетом расстояния между буквами
20
  image = Image.new('L', (base_width, height), 'black')
21
  draw = ImageDraw.Draw(image)
22
  text_width = 0
23
  for char in text:
24
  text_bbox = draw.textbbox((0, 0), char, font=font)
25
  text_width += text_bbox[2] - text_bbox[0] + letter_spacing
26
- text_width -= letter_spacing # Убираем дополнительный интервал после последней буквы
27
 
28
- # Увеличиваем ширину изображения, если текст не помещается
29
  if text_width + margin * 2 > base_width:
30
  width = text_width + margin * 2
31
  else:
32
  width = base_width
33
 
34
- # Создаем изображение с новой шириной
35
  image = Image.new('L', (width, height), 'black')
36
  draw = ImageDraw.Draw(image)
37
 
38
- # Пишем текст в центре изображения
39
  text_x = (width - text_width) // 2
40
  text_y = (height - (text_bbox[3] - text_bbox[1])) // 2
41
  for char in text:
@@ -43,32 +43,32 @@ def text_to_spectrogram_image(text, base_width=512, height=256, max_font_size=80
43
  char_bbox = draw.textbbox((0, 0), char, font=font)
44
  text_x += char_bbox[2] - char_bbox[0] + letter_spacing
45
 
46
- # Повышаем контрастность текста
47
  image = np.array(image)
48
- image = np.where(image > 0, 255, image) # Устанавливаем текст как максимально белый
49
  return image
50
 
51
- # Преобразовываем изображение в аудиосигнал
52
  def spectrogram_image_to_audio(image, sr=22050):
53
- # Переворачиваем изображение по вертикали
54
  flipped_image = np.flipud(image)
55
 
56
  # Преобразуем изображение в амплитуды спектрограммы
57
  S = flipped_image.astype(np.float32) / 255.0 * 100.0
58
 
59
- # Преобразуем спектрограмму в аудиосигнал
60
  y = librosa.griffinlim(S)
61
  return y
62
 
63
- # Функция для создания аудиофайла и спектрограммы из текста
64
  def create_audio_with_spectrogram(text, base_width, height, max_font_size, margin, letter_spacing):
65
- # Создаем изображение спектрограммы с нормальным текстом
66
  spec_image = text_to_spectrogram_image(text, base_width, height, max_font_size, margin, letter_spacing)
67
 
68
- # Генерируем аудиосигнал с перевернутым текстом
69
  y = spectrogram_image_to_audio(spec_image)
70
 
71
- # Сохраняем аудиосигнал и изображение спектрограммы
72
  audio_path = 'output.wav'
73
  sf.write(audio_path, y, 22050)
74
 
@@ -77,35 +77,35 @@ def create_audio_with_spectrogram(text, base_width, height, max_font_size, margi
77
 
78
  return audio_path, image_path
79
 
80
- # Интерфейс Gradio
81
- with gr.Blocks(title='Аудио-Стеганография', theme=gr.themes.Soft(primary_hue="green", secondary_hue="green", spacing_size="sm", radius_size="lg")) as iface:
82
 
83
  with gr.Group():
84
  with gr.Row(variant='panel'):
85
  with gr.Column():
86
- gr.HTML("<center><h2><a href='https://t.me/pol1trees'>Telegram Канал</a></h2></center>")
87
  with gr.Column():
88
- gr.HTML("<center><h2><a href='https://t.me/+GMTP7hZqY0E4OGRi'>Telegram Чат</a></h2></center>")
89
  with gr.Column():
90
  gr.HTML("<center><h2><a href='https://www.youtube.com/channel/UCHb3fZEVxUisnqLqCrEM8ZA'>YouTube</a></h2></center>")
91
  with gr.Column():
92
  gr.HTML("<center><h2><a href='https://github.com/Bebra777228/Audio-Steganography'>GitHub</a></h2></center>")
93
 
94
  with gr.Group():
95
- text = gr.Textbox(lines=2, placeholder="Введите свой текст:", label="Текст")
96
 
97
  with gr.Row(variant='panel'):
98
- base_width = gr.Slider(value=512, label="Ширина изображения", visible=False)
99
- height = gr.Slider(value=256, label="Высота изображения", visible=False)
100
- max_font_size = gr.Slider(minimum=10, maximum=130, step=5, value=80, label="Размер шрифта")
101
- margin = gr.Slider(minimum=0, maximum=50, step=1, value=10, label="Отступ")
102
- letter_spacing = gr.Slider(minimum=0, maximum=50, step=1, value=5, label="Расстояние между буквами")
103
 
104
- generate_button = gr.Button("Сгенерировать")
105
 
106
  with gr.Column(variant='panel'):
107
- output_audio = gr.Audio(type="filepath", label="Сгенерированный звук")
108
- output_image = gr.Image(type="filepath", label="Спектрограмма")
109
 
110
  def gradio_interface_fn(text, base_width, height, max_font_size, margin, letter_spacing):
111
  print("\n", text)
 
7
  import soundfile as sf
8
  import os
9
 
10
+ # Function for creating a spectrogram image with text
11
  def text_to_spectrogram_image(text, base_width=512, height=256, max_font_size=80, margin=10, letter_spacing=5):
12
+ # Font and text size
13
  font_path = "/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf"
14
  if os.path.exists(font_path):
15
  font = ImageFont.truetype(font_path, max_font_size)
16
  else:
17
  font = ImageFont.load_default()
18
 
19
+ # Determine the width of the text, taking into account the distance between letters
20
  image = Image.new('L', (base_width, height), 'black')
21
  draw = ImageDraw.Draw(image)
22
  text_width = 0
23
  for char in text:
24
  text_bbox = draw.textbbox((0, 0), char, font=font)
25
  text_width += text_bbox[2] - text_bbox[0] + letter_spacing
26
+ text_width -= letter_spacing # Remove the extra spacing after the last letter
27
 
28
+ # Increase the width of the image if the text does not fit
29
  if text_width + margin * 2 > base_width:
30
  width = text_width + margin * 2
31
  else:
32
  width = base_width
33
 
34
+ # Create an image with a new width
35
  image = Image.new('L', (width, height), 'black')
36
  draw = ImageDraw.Draw(image)
37
 
38
+ # Writing text in the center of the image
39
  text_x = (width - text_width) // 2
40
  text_y = (height - (text_bbox[3] - text_bbox[1])) // 2
41
  for char in text:
 
43
  char_bbox = draw.textbbox((0, 0), char, font=font)
44
  text_x += char_bbox[2] - char_bbox[0] + letter_spacing
45
 
46
+ # Increase text contrast
47
  image = np.array(image)
48
+ image = np.where(image > 0, 255, image) # Setting the text to "maximum white"
49
  return image
50
 
51
+ # Converting an image to audio
52
  def spectrogram_image_to_audio(image, sr=22050):
53
+ # Rotate the image vertically
54
  flipped_image = np.flipud(image)
55
 
56
  # Преобразуем изображение в амплитуды спектрограммы
57
  S = flipped_image.astype(np.float32) / 255.0 * 100.0
58
 
59
+ # Converting the spectrogram to an audio signal
60
  y = librosa.griffinlim(S)
61
  return y
62
 
63
+ # Function for creating an audio file and spectrogram from text
64
  def create_audio_with_spectrogram(text, base_width, height, max_font_size, margin, letter_spacing):
65
+ # Create spectrogram image with normal text
66
  spec_image = text_to_spectrogram_image(text, base_width, height, max_font_size, margin, letter_spacing)
67
 
68
+ # Generate audio signal with inverted text
69
  y = spectrogram_image_to_audio(spec_image)
70
 
71
+ # Save audio signal and spectrogram image
72
  audio_path = 'output.wav'
73
  sf.write(audio_path, y, 22050)
74
 
 
77
 
78
  return audio_path, image_path
79
 
80
+ # Gradio interface
81
+ with gr.Blocks(title='Audio Steganography', theme=gr.themes.Soft(primary_hue="green", secondary_hue="green", spacing_size="sm", radius_size="lg")) as iface:
82
 
83
  with gr.Group():
84
  with gr.Row(variant='panel'):
85
  with gr.Column():
86
+ gr.HTML("<center><h2><a href='https://t.me/pol1trees'>Telegram Channel</a></h2></center>")
87
  with gr.Column():
88
+ gr.HTML("<center><h2><a href='https://t.me/+GMTP7hZqY0E4OGRi'>Telegram Chat</a></h2></center>")
89
  with gr.Column():
90
  gr.HTML("<center><h2><a href='https://www.youtube.com/channel/UCHb3fZEVxUisnqLqCrEM8ZA'>YouTube</a></h2></center>")
91
  with gr.Column():
92
  gr.HTML("<center><h2><a href='https://github.com/Bebra777228/Audio-Steganography'>GitHub</a></h2></center>")
93
 
94
  with gr.Group():
95
+ text = gr.Textbox(lines=2, placeholder="Enter your text:", label="Текст")
96
 
97
  with gr.Row(variant='panel'):
98
+ base_width = gr.Slider(value=512, label="Image Width", visible=False)
99
+ height = gr.Slider(value=256, label="Image Height", visible=False)
100
+ max_font_size = gr.Slider(minimum=10, maximum=130, step=5, value=80, label="Font Size")
101
+ margin = gr.Slider(minimum=0, maximum=50, step=1, value=10, label="Indent")
102
+ letter_spacing = gr.Slider(minimum=0, maximum=50, step=1, value=5, label="Letter spacing")
103
 
104
+ generate_button = gr.Button("Generate")
105
 
106
  with gr.Column(variant='panel'):
107
+ output_audio = gr.Audio(type="filepath", label="Generated audio")
108
+ output_image = gr.Image(type="filepath", label="Spectrogram")
109
 
110
  def gradio_interface_fn(text, base_width, height, max_font_size, margin, letter_spacing):
111
  print("\n", text)