Update app.py
Browse files
app.py
CHANGED
|
@@ -22,6 +22,56 @@ def home():
|
|
| 22 |
return "Flask Video Generator is Running"
|
| 23 |
|
| 24 |
@app.route("/generate", methods=["POST"])
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
def generate_video():
|
| 26 |
try:
|
| 27 |
data = request.get_json()
|
|
@@ -91,56 +141,6 @@ def generate_video():
|
|
| 91 |
image_olst=[]
|
| 92 |
for id in range(len(lines)):
|
| 93 |
create_text_image(lines[id],id,image_olst)
|
| 94 |
-
def video_func(id):
|
| 95 |
-
tts = gTTS(text=lines[id], lang='ta', slow=False)
|
| 96 |
-
filename = "audio"+str(id)+".mp3"
|
| 97 |
-
tts.save(filename)
|
| 98 |
-
if os.path.exists(filename):
|
| 99 |
-
audio = MP3(filename)
|
| 100 |
-
duration = audio.info.length
|
| 101 |
-
IMAGE_PATH = "images/slide"+str(id)+".png" # Ensure this path is correct
|
| 102 |
-
VIDEO_DURATION = duration # seconds
|
| 103 |
-
HIGHLIGHT_COLOR = (255, 255, 0) # Yellow highlight
|
| 104 |
-
HIGHLIGHT_OPACITY = 0.5 # Semi-transparent
|
| 105 |
-
|
| 106 |
-
# --- OCR STEP ---
|
| 107 |
-
img = Image.open(IMAGE_PATH)
|
| 108 |
-
data = pytesseract.image_to_data(img, output_type=pytesseract.Output.DICT)
|
| 109 |
-
|
| 110 |
-
# Extract words and their positions
|
| 111 |
-
words = []
|
| 112 |
-
for i in range(len(data['text'])):
|
| 113 |
-
word = data['text'][i].strip()
|
| 114 |
-
if word and int(data['conf'][i]) > 60:
|
| 115 |
-
x, y, w, h = data['left'][i], data['top'][i], data['width'][i], data['height'][i]
|
| 116 |
-
words.append({'text': word, 'box': (x, y, w, h)})
|
| 117 |
-
|
| 118 |
-
# --- BASE IMAGE CLIP ---
|
| 119 |
-
image_clip = ImageClip(IMAGE_PATH).set_duration(VIDEO_DURATION)
|
| 120 |
-
|
| 121 |
-
# --- HIGHLIGHT WORDS ONE BY ONE ---
|
| 122 |
-
n_words = len(words)
|
| 123 |
-
highlight_duration = VIDEO_DURATION / n_words
|
| 124 |
-
|
| 125 |
-
highlight_clips = []
|
| 126 |
-
|
| 127 |
-
for i, word in enumerate(words):
|
| 128 |
-
x, y, w, h = word['box']
|
| 129 |
-
start = i * highlight_duration
|
| 130 |
-
end = start + highlight_duration
|
| 131 |
-
|
| 132 |
-
# Create highlight rectangle
|
| 133 |
-
rect = ColorClip(size=(w, h), color=HIGHLIGHT_COLOR)
|
| 134 |
-
rect = rect.set_opacity(HIGHLIGHT_OPACITY).set_position((x, y)).set_start(start).set_end(end)
|
| 135 |
-
|
| 136 |
-
highlight_clips.append(rect)
|
| 137 |
-
|
| 138 |
-
# --- FINAL VIDEO --
|
| 139 |
-
|
| 140 |
-
final_clip = CompositeVideoClip([image_clip] + highlight_clips)
|
| 141 |
-
audio = AudioFileClip(filename)
|
| 142 |
-
final_clip = final_clip.set_audio(audio)
|
| 143 |
-
final_clip.write_videofile("clip"+str(id)+".mp4", fps=24)
|
| 144 |
id=list(range(len(lines)))
|
| 145 |
with Pool(processes=cpu_count()) as pool:
|
| 146 |
pool.map(video_func,id)
|
|
|
|
| 22 |
return "Flask Video Generator is Running"
|
| 23 |
|
| 24 |
@app.route("/generate", methods=["POST"])
|
| 25 |
+
def video_func(id):
|
| 26 |
+
tts = gTTS(text=lines[id], lang='ta', slow=False)
|
| 27 |
+
filename = "audio"+str(id)+".mp3"
|
| 28 |
+
tts.save(filename)
|
| 29 |
+
if os.path.exists(filename):
|
| 30 |
+
audio = MP3(filename)
|
| 31 |
+
duration = audio.info.length
|
| 32 |
+
IMAGE_PATH = "images/slide"+str(id)+".png" # Ensure this path is correct
|
| 33 |
+
VIDEO_DURATION = duration # seconds
|
| 34 |
+
HIGHLIGHT_COLOR = (255, 255, 0) # Yellow highlight
|
| 35 |
+
HIGHLIGHT_OPACITY = 0.5 # Semi-transparent
|
| 36 |
+
|
| 37 |
+
# --- OCR STEP ---
|
| 38 |
+
img = Image.open(IMAGE_PATH)
|
| 39 |
+
data = pytesseract.image_to_data(img, output_type=pytesseract.Output.DICT)
|
| 40 |
+
|
| 41 |
+
# Extract words and their positions
|
| 42 |
+
words = []
|
| 43 |
+
for i in range(len(data['text'])):
|
| 44 |
+
word = data['text'][i].strip()
|
| 45 |
+
if word and int(data['conf'][i]) > 60:
|
| 46 |
+
x, y, w, h = data['left'][i], data['top'][i], data['width'][i], data['height'][i]
|
| 47 |
+
words.append({'text': word, 'box': (x, y, w, h)})
|
| 48 |
+
|
| 49 |
+
# --- BASE IMAGE CLIP ---
|
| 50 |
+
image_clip = ImageClip(IMAGE_PATH).set_duration(VIDEO_DURATION)
|
| 51 |
+
|
| 52 |
+
# --- HIGHLIGHT WORDS ONE BY ONE ---
|
| 53 |
+
n_words = len(words)
|
| 54 |
+
highlight_duration = VIDEO_DURATION / n_words
|
| 55 |
+
|
| 56 |
+
highlight_clips = []
|
| 57 |
+
|
| 58 |
+
for i, word in enumerate(words):
|
| 59 |
+
x, y, w, h = word['box']
|
| 60 |
+
start = i * highlight_duration
|
| 61 |
+
end = start + highlight_duration
|
| 62 |
+
|
| 63 |
+
# Create highlight rectangle
|
| 64 |
+
rect = ColorClip(size=(w, h), color=HIGHLIGHT_COLOR)
|
| 65 |
+
rect = rect.set_opacity(HIGHLIGHT_OPACITY).set_position((x, y)).set_start(start).set_end(end)
|
| 66 |
+
|
| 67 |
+
highlight_clips.append(rect)
|
| 68 |
+
|
| 69 |
+
# --- FINAL VIDEO --
|
| 70 |
+
|
| 71 |
+
final_clip = CompositeVideoClip([image_clip] + highlight_clips)
|
| 72 |
+
audio = AudioFileClip(filename)
|
| 73 |
+
final_clip = final_clip.set_audio(audio)
|
| 74 |
+
final_clip.write_videofile("clip"+str(id)+".mp4", fps=24)
|
| 75 |
def generate_video():
|
| 76 |
try:
|
| 77 |
data = request.get_json()
|
|
|
|
| 141 |
image_olst=[]
|
| 142 |
for id in range(len(lines)):
|
| 143 |
create_text_image(lines[id],id,image_olst)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 144 |
id=list(range(len(lines)))
|
| 145 |
with Pool(processes=cpu_count()) as pool:
|
| 146 |
pool.map(video_func,id)
|