Rtechs_utilities / media_utils.py
RO-Rtechs's picture
Update media_utils.py
a53b85e verified
raw
history blame
No virus
18.4 kB
import gradio as gr
import subprocess
import os
from tqdm import tqdm
import time
from PIL import Image
from concurrent.futures import ThreadPoolExecutor
import torch
def get_device():
if torch.cuda.is_available():
return torch.device('cuda')
else:
return torch.device('cpu')
device = get_device()
print(f"Using device: {device}")
def get_unique_filename(base_name, extension):
counter = 1
unique_name = f"{base_name}{extension}"
while os.path.exists(unique_name):
unique_name = f"{base_name}_{counter}{extension}"
counter += 1
return unique_name
def combine_videos(video_files):
if not video_files:
return "No video files selected.", None
output_file = get_unique_filename("combined_video", ".mp4")
filelist_path = os.path.abspath("filelist.txt")
with open(filelist_path, "w") as filelist:
for video in video_files:
video_path = os.path.abspath(video).replace('\\', '/')
filelist.write(f"file '{video_path}'\n")
command = [
"ffmpeg", "-f", "concat", "-safe", "0", "-i", filelist_path,
"-c", "copy", output_file
]
with ThreadPoolExecutor() as executor:
future = executor.submit(subprocess.run, command, text=True, capture_output=True)
result = future.result() # Waits for the command to complete and returns the result
if result.returncode == 0:
return f"Videos combined successfully into {output_file}", output_file
else:
return f"Error combining videos: {result.stderr}", None
def combine_audios(audio_files):
if not audio_files:
return "No audio files selected.", None
output_file = get_unique_filename("combined_audio", ".mp3")
filelist_path = os.path.abspath("filelist.txt")
with open(filelist_path, "w") as filelist:
for audio in audio_files:
filelist.write(f"file '{os.path.abspath(audio).replace('\\', '/')}'\n")
command = [
"ffmpeg", "-f", "concat", "-safe", "0", "-i", filelist_path,
"-c", "copy", output_file
]
with ThreadPoolExecutor() as executor:
future = executor.submit(subprocess.run, command, text=True, capture_output=True)
result = future.result() # Waits for the command to complete and returns the result
if result.returncode == 0:
return f"Audios combined successfully into {output_file}", output_file
else:
return f"Error combining audios: {result.stderr}", None
def combine_images(image_files):
if not image_files:
return "No image files selected.", None
output_file = get_unique_filename("combined_image", ".mp4")
command = ["convert"] + image_files + [output_file]
process = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
total_time = 0
with tqdm(total=100, desc="Combining Images") as pbar:
while True:
output = process.stderr.readline()
if output == '' and process.poll() is not None:
break
if output:
total_time += 1
pbar.update(1)
time.sleep(0.1)
return f"Images combined successfully into {output_file}", output_file
def adjust_speed(media_file, speed):
if not media_file:
return "No media file selected.", None
output_file = get_unique_filename(f"adjusted_speed_{os.path.basename(media_file)}", ".mp4")
command = [
"ffmpeg", "-i", media_file, "-filter:v", f"setpts={1/speed}*PTS",
"-filter:a", f"atempo={speed}", output_file
]
process = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
total_time = 0
with tqdm(total=100, desc="Adjusting Speed") as pbar:
while True:
output = process.stderr.readline()
if output == '' and process.poll() is not None:
break
if output:
total_time += 1
pbar.update(1)
time.sleep(0.1)
return f"Speed adjusted successfully to {speed}x in {output_file}", output_file
def adjust_speed_by_length(media_file, desired_length_hhmmss):
if not media_file:
return "No media file selected.", None
original_length = get_media_length(media_file)
desired_length = hhmmss_to_seconds(desired_length_hhmmss)
speed = original_length / desired_length
return adjust_speed(media_file, speed)
def adjust_speed_combined(media_file, speed, hours, minutes, seconds, compress):
if hours or minutes or seconds:
desired_length = f"{int(hours):02}:{int(minutes):02}:{int(seconds):02}"
output_message, output_file = adjust_speed_by_length(media_file, desired_length)
else:
output_message, output_file = adjust_speed(media_file, speed)
if compress and output_file:
compressed_output_file = get_unique_filename(f"compressed_{os.path.basename(output_file)}", ".mp4")
compress_video(output_file, compressed_output_file)
output_message = f"{output_message} and compressed to {compressed_output_file}"
output_file = compressed_output_file
return output_message, output_file
def hhmmss_to_seconds(hhmmss):
h, m, s = map(int, hhmmss.split(':'))
return h * 3600 + m * 60 + s
def get_media_length(media_file):
result = subprocess.run(
["ffprobe", "-v", "error", "-show_entries", "format=duration", "-of", "default=noprint_wrappers=1:nokey=1", media_file],
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT
)
return float(result.stdout)
def auto_color_correct(video_file):
if not video_file:
return "No video file selected.", None
output_file = get_unique_filename(f"color_corrected_{os.path.basename(video_file)}", "")
command = [
"ffmpeg", "-i", video_file, "-vf", "eq=brightness=0.06:saturation=1.5", output_file
]
subprocess.run(command, check=True)
return f"Auto color correction applied successfully to {output_file}", output_file
def extract_audio(video_file):
if not video_file:
return "No video file selected.", None
output_file = get_unique_filename(f"extracted_audio_{os.path.basename(video_file)}", ".mka")
command = [
"ffmpeg", "-i", video_file, "-vn", "-acodec", "copy", output_file
]
try:
subprocess.run(command, check=True)
except subprocess.CalledProcessError as e:
return f"Error extracting audio: {e}. Command: {' '.join(command)}", None
return f"Audio extracted successfully into {output_file}", output_file
def compress_video(input_file, output_file):
command = [
"ffmpeg", "-i", input_file, "-vcodec", "libx265", "-crf", "28", output_file
]
subprocess.run(command, check=True)
return output_file
def add_watermark(video_file, watermark_type, watermark_text, watermark_image, opacity, position_x, position_y, font_size, font_color, resize_width, resize_height):
if not video_file:
return "No video file selected.", None
output_file = get_unique_filename(f"watermarked_{os.path.basename(video_file)}", ".mp4")
drawtext = f"drawtext=text='{watermark_text}':x={position_x}:y={position_y}:fontsize={font_size}:fontcolor={font_color}@{opacity}" if watermark_type == "text" else ""
def add_watermark(video_file, watermark_type, watermark_text, watermark_image, opacity, position, font_size, font_color, resize):
if not video_file:
return "No video file selected.", None
output_file = get_unique_filename(f"watermarked_{os.path.basename(video_file)}", ".mp4")
drawtext = f"drawtext=text='{watermark_text}':x={position[0]}:y={position[1]}:fontsize={font_size}:fontcolor={font_color}@{opacity}" if watermark_type == "text" else ""
overlay = f"overlay={position[0]}:{position[1]}" if watermark_type == "image" else ""
resize_cmd = f"scale={resize[0]}:{resize[1]}" if resize else "scale=iw:ih"
command = [
"ffmpeg", "-i", video_file,
"-vf", f"{resize_cmd},{drawtext if watermark_type == 'text' else ''}{overlay if watermark_type == 'image' else ''}",
output_file
]
process = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
total_time = 0
with tqdm(total=100, desc="Adding Watermark") as pbar:
while True:
output = process.stderr.readline()
if output == '' and process.poll() is not None:
break
if output:
total_time += 1
pbar.update(1)
time.sleep(0.1)
return f"Watermark added successfully to {output_file}", output_file
def compress_image(input_image):
if not input_image:
return "No image file selected.", None
output_file = get_unique_filename(f"compressed_{os.path.basename(input_image)}", ".jpg")
with Image.open(input_image) as img:
img.save(output_file, "JPEG", quality=95) # Adjust quality as needed
return f"Image compressed successfully into {output_file}", output_file
def compress_image_lossless(input_image):
if not input_image:
return "No image file selected.", None
output_file = get_unique_filename(f"compressed_lossless_{os.path.basename(input_image)}", ".png")
with Image.open(input_image) as img:
img.save(output_file, "PNG", optimize=True) # PNG is lossless
return f"Image compressed losslessly into {output_file}", output_file
def mp3_to_video(mp3_file, image_file):
if not mp3_file or not image_file:
return "MP3 file or image file not selected.", None
output_file = get_unique_filename(f"{os.path.splitext(os.path.basename(mp3_file))[0]}", ".mp4")
command = [
"ffmpeg", "-loop", "1", "-i", image_file, "-i", mp3_file, "-c:v", "libx264", "-c:a", "aac", "-b:a", "192k", "-shortest", output_file
]
subprocess.run(command, check=True)
return f"MP3 converted to video successfully into {output_file}", output_file
def video_to_mp3(video_file):
if not video_file:
return "No video file selected.", None
output_file = get_unique_filename(f"{os.path.splitext(os.path.basename(video_file))[0]}", ".mp3")
command = [
"ffmpeg", "-i", video_file, "-q:a", "0", "-map", "a", output_file
]
subprocess.run(command, check=True)
return f"Video converted to MP3 successfully into {output_file}", output_file
def convert_image_format(input_image, output_format):
if not input_image:
return "No image file selected.", None
output_file = get_unique_filename(f"{os.path.splitext(os.path.basename(input_image))[0]}", f".{output_format}")
with Image.open(input_image) as img:
img.save(output_file, output_format.upper())
return f"Image converted to {output_format} format successfully into {output_file}", output_file
def interface():
with gr.Blocks(theme="small_and_pretty") as demo: # Applying the "small_and_pretty" theme
gr.Markdown("### Media Combiner Tool")
with gr.Tab("Combine Videos"):
video_files = gr.File(label="Select Video Files", type="filepath", file_count="multiple")
video_submit = gr.Button("Combine Videos")
video_output = gr.Textbox(label="Output")
video_download = gr.File(label="Download Combined Video")
video_submit.click(combine_videos, inputs=[video_files], outputs=[video_output, video_download])
with gr.Tab("Combine Audios"):
audio_files = gr.File(label="Select Audio Files", type="filepath", file_count="multiple")
audio_submit = gr.Button("Combine Audios")
audio_output = gr.Textbox(label="Output")
audio_download = gr.File(label="Download Combined Audio")
audio_submit.click(combine_audios, inputs=[audio_files], outputs=[audio_output, audio_download])
with gr.Tab("Combine Images"):
image_files = gr.File(label="Select Image Files", type="filepath", file_count="multiple")
image_submit = gr.Button("Combine Images")
image_output = gr.Textbox(label="Output")
image_download = gr.File(label="Download Combined Image")
image_submit.click(combine_images, inputs=[image_files], outputs=[image_output, image_download])
with gr.Tab("Adjust Speed"):
media_file = gr.File(label="Select Media File", type="filepath")
speed = gr.Slider(label="Speed", minimum=0.5, maximum=2.0, step=0.1, value=1.0)
with gr.Row():
hours = gr.Number(label="Hours", value=0, precision=0)
minutes = gr.Number(label="Minutes", value=0, precision=0)
seconds = gr.Number(label="Seconds", value=0, precision=0)
compress = gr.Checkbox(label="Compress Video", value=False)
speed_submit = gr.Button("Adjust Speed")
speed_output = gr.Textbox(label="Output")
speed_download = gr.File(label="Download Adjusted Media")
speed_submit.click(adjust_speed_combined, inputs=[media_file, speed, hours, minutes, seconds, compress], outputs=[speed_output, speed_download])
with gr.Tab("Auto Color Correction"):
video_file = gr.File(label="Select Video File", type="filepath")
color_submit = gr.Button("Apply Color Correction")
color_output = gr.Textbox(label="Output")
color_download = gr.File(label="Download Color Corrected Video")
color_submit.click(auto_color_correct, inputs=[video_file], outputs=[color_output, color_download])
with gr.Tab("Extract Audio"):
video_file = gr.File(label="Select Video File", type="filepath")
extract_submit = gr.Button("Extract Audio")
extract_output = gr.Textbox(label="Output")
extract_download = gr.File(label="Download Extracted Audio")
extract_submit.click(extract_audio, inputs=[video_file], outputs=[extract_output, extract_download])
with gr.Tab("Add Watermark"):
video_file = gr.File(label="Select Video File", type="filepath")
watermark_type = gr.Radio(label="Watermark Type", choices=["text", "image"], value="text")
watermark_text = gr.Textbox(label="Watermark Text", visible=True)
watermark_image = gr.File(label="Watermark Image", type="filepath", visible=False)
opacity = gr.Slider(label="Opacity", minimum=0.0, maximum=1.0, step=0.1, value=1.0)
position_x = gr.Slider(label="Position X", minimum=0, maximum=1920, step=1, value=0)
position_y = gr.Slider(label="Position Y", minimum=0, maximum=1080, step=1, value=0)
font_size = gr.Slider(label="Font Size", minimum=10, maximum=100, step=1, value=24, visible=True)
font_color = gr.ColorPicker(label="Font Color", value="#FFFFFF", visible=True)
resize_width = gr.Number(label="Resize Width", visible=False)
resize_height = gr.Number(label="Resize Height", visible=False)
watermark_submit = gr.Button("Add Watermark")
watermark_output = gr.Textbox(label="Output")
watermark_download = gr.File(label="Download Watermarked Video")
def update_visibility(watermark_type):
return {
watermark_text: gr.update(visible=watermark_type == "text"),
watermark_image: gr.update(visible=watermark_type == "image"),
font_size: gr.update(visible=watermark_type == "text"),
font_color: gr.update(visible=watermark_type == "text"),
resize_width: gr.update(visible=watermark_type == "image"),
resize_height: gr.update(visible=watermark_type == "image")
}
watermark_type.change(update_visibility, inputs=[watermark_type], outputs=[watermark_text, watermark_image, font_size, font_color, resize_width, resize_height])
watermark_submit.click(add_watermark, inputs=[video_file, watermark_type, watermark_text, watermark_image, opacity, position_x, position_y, font_size, font_color, resize_width, resize_height], outputs=[watermark_output, watermark_download])
with gr.Tab("Compress Image Losslessly"):
image_file = gr.File(label="Select Image File", type="filepath")
compress_submit = gr.Button("Compress Image Losslessly")
compress_output = gr.Textbox(label="Output")
compress_download = gr.File(label="Download Compressed Image")
compress_submit.click(compress_image_lossless, inputs=[image_file], outputs=[compress_output, compress_download])
with gr.Tab("Convert MP3 to Video"):
mp3_file = gr.File(label="Select MP3 File", type="filepath")
image_file = gr.File(label="Select Image File", type="filepath")
convert_submit = gr.Button("Convert MP3 to Video")
convert_output = gr.Textbox(label="Output")
convert_download = gr.File(label="Download Converted Video")
convert_submit.click(mp3_to_video, inputs=[mp3_file, image_file], outputs=[convert_output, convert_download])
with gr.Tab("Convert Video to MP3"):
video_file = gr.File(label="Select Video File", type="filepath")
convert_submit = gr.Button("Convert Video to MP3")
convert_output = gr.Textbox(label="Output")
convert_download = gr.File(label="Download Converted MP3")
convert_submit.click(video_to_mp3, inputs=[video_file], outputs=[convert_output, convert_download])
with gr.Tab("Convert Image Format"):
input_image = gr.File(label="Select Image File", type="filepath")
output_format = gr.Dropdown(label="Output Format", choices=["png", "jpg", "bmp", "gif"], value="png")
convert_submit = gr.Button("Convert Image Format")
convert_output = gr.Textbox(label="Output")
convert_download = gr.File(label="Download Converted Image")
convert_submit.click(convert_image_format, inputs=[input_image, output_format], outputs=[convert_output, convert_download])
return demo
if __name__ == "__main__":
demo = interface()
demo.launch(share=True)