fmusvid 0.0.1

A Python video processing library

FMUS-VID Library

FMUS-VID is a Python library for video processing and manipulation.

Installation

# Basic installation
pip install fmusvid

# With AI features
pip install fmusvid[ai]

# With streaming support
pip install fmusvid[streaming]

# With GUI support
pip install fmusvid[gui]

# With all features
pip install fmusvid[all]

Quick Start

Basic Video Processing

from fmusvid.video import VideoProcessor
from fmusvid.operations import basic

# Load a video
video = VideoProcessor("input.mp4")

# Apply operations
video.apply(basic.cut(start=10, end=20))
video.apply(basic.resize(width=1280, height=720))

# Save the result
video.save("output.mp4")

Object Detection

from fmusvid.video import VideoProcessor
from fmusvid.ai import detection

# Load a video
video = VideoProcessor("input.mp4")

# Detect objects
results = video.apply(detection.detect_objects(model="yolo"))

# Save the result with bounding boxes
video.save("output_with_detection.mp4")

RTMP Streaming with Multilingual Subtitles

from fmusvid.stream import RTMPCapture, RTMPOutput
from fmusvid.ai.speech import WhisperTranscriber
from fmusvid.ai.translation import LLMTranslator
from fmusvid.operations.text import SubtitleOverlay

# Create components
rtmp_capture = RTMPCapture("rtmp://localhost/live/stream")
transcriber = WhisperTranscriber(model="medium", device="cuda")
translator = LLMTranslator(model="ollama:mistral", target_languages=["es", "fr", "de"])

# Create outputs for different languages
outputs = {
    "en": RTMPOutput("rtmp://localhost/live/stream_en"),
    "es": RTMPOutput("rtmp://localhost/live/stream_es"),
    "fr": RTMPOutput("rtmp://localhost/live/stream_fr"),
    "de": RTMPOutput("rtmp://localhost/live/stream_de")
}

# Process frames in real-time
async for frame, audio in rtmp_capture.aiter_frames_with_audio():
    # Transcribe audio
    text = await transcriber.transcribe_chunk(audio)

    # Translate text
    translations = await translator.translate_async(text)

    # Create frames with different subtitles
    frames = {}
    for lang, translation in translations.items():
        subtitle = SubtitleOverlay(position="bottom")
        frames[lang] = subtitle.apply(frame, translation)

    # Write frames to outputs
    for lang, output_frame in frames.items():
        await outputs[lang].write_frame(output_frame)

Webcam Capture

from fmusvid.capture import CameraCapture

# List available cameras
from fmusvid.capture import list_cameras
cameras = list_cameras()
for camera in cameras:
    print(f"Camera {camera['id']}: {camera['name']} ({camera['width']}x{camera['height']})")

# Create a camera capture instance
camera = CameraCapture(camera_id=0, width=1280, height=720)

# Start the camera
camera.start()

# Capture a frame
frame = camera.get_frame()

# Capture an image and save it
camera.capture_image("webcam_capture.jpg")

# Record video
camera.record_video("webcam_recording.mp4", duration=10)  # 10 seconds

# Stop the camera
camera.stop()

GUI Applications

FMUS-VID includes a PyQt6-based GUI application for webcam capture:

# Install with GUI support
pip install fmusvid[gui]

# Run the camera GUI application
python examples/camera_gui.py

Example Scripts

The library includes several example scripts to demonstrate its capabilities:

• examples/basic_example.py: Basic video processing operations
• examples/transitions_example.py: Video transitions and effects
• examples/ai_features.py: Object detection and tracking
• examples/rtmp_live_example.py: RTMP streaming with real-time transcription and translation
• examples/rtmp_demo.py: Demo of RTMP capabilities using local video files
• examples/camera_gui.py: PyQt6-based webcam capture application

Command Line Interface

FMUS-VID provides a comprehensive command-line interface for common video operations:

# Basic CLI usage
fmusvid [command] [options]

Available Commands

Command	Description	Example
convert	Convert video format	fmusvid convert input.mp4 output.webm
cut	Cut a segment from video	fmusvid cut -i input.mp4 -o output.mp4 -s 10 -e 20
resize	Resize video	fmusvid resize -i input.mp4 -o output.mp4 -w 1280 -h 720
info	Display video information	fmusvid info input.mp4
concat	Concatenate multiple videos	fmusvid concat -i video1.mp4 video2.mp4 -o output.mp4
extract-audio	Extract audio from video	fmusvid extract-audio -i input.mp4 -o audio.mp3
add-subtitles	Add subtitles to video	fmusvid add-subtitles -i input.mp4 -s subtitles.srt -o output.mp4
generate-subtitles	Generate subtitles using AI	fmusvid generate-subtitles -i input.mp4 -o subtitles.srt --model medium
translate-subtitles	Translate subtitles	fmusvid translate-subtitles -i subtitles.srt -o subtitles_es.srt --target es
camera	Capture from webcam	fmusvid camera -o capture.mp4 -d 10
screen	Capture screen recording	fmusvid screen -o screen.mp4 -d 30
rtmp-stream	Stream video to RTMP server	fmusvid rtmp-stream -i input.mp4 -u rtmp://server/live/stream
rtmp-capture	Capture RTMP stream	fmusvid rtmp-capture -u rtmp://server/live/stream -o output.mp4 -d 60

Global Options

Option	Description
-v, --verbose	Enable verbose output
--version	Show version information
--help	Show help message
--config FILE	Use custom configuration file
--log-level LEVEL	Set logging level (debug, info, warning, error)

For more detailed help on each command:

fmusvid [command] --help

Dependencies

• OpenCV
• NumPy
• FFmpeg
• PyTorch (for AI features)
• Faster-Whisper (for speech recognition)
• Local LLM support (for translation)
• PyQt6 (for GUI applications)

License

This project is licensed under the MIT License.