Skip to content

raqueladaia/video_processor

BranchesTags

Repository files navigation

Video Processing Tools

A comprehensive toolkit for video processing with five main capabilities:

  1. Long Video Chopping - Segment videos into smaller consecutive chunks
  2. Snippet Selection - Extract specific segments around timestamps of interest
  3. Adjust Brightness - GUI tool for adjusting video brightness and contrast
  4. Crop Video - GUI tool for cropping videos into multiple regions
  5. Video Metadata Check - Extract and analyze comprehensive video metadata

Features

Long Video Chopping

  • Segments long videos into smaller chunks of specified duration
  • Supports single video files or entire folders
  • Maintains original video quality
  • Never modifies original files
  • Creates organized output folder structure

Snippet Selection

  • Extracts video snippets around specific timestamps using FFmpeg
  • Reads timestamp data from Excel files (supports multiple sheets)
  • Configurable duration before/after each timestamp
  • Supports videos in subdirectories
  • Automatic animal ID extraction from video names
  • Pain/non-pain classification support
  • Generates CSV reports with metadata
  • Handles missing videos gracefully
  • Fast, lossless video extraction

Video Metadata Check

  • Extracts comprehensive video metadata (FPS, resolution, duration, etc.)
  • Compares multiple videos for consistency
  • Validates videos against user-specified criteria
  • Detects anomalies and outliers
  • Generates reports in multiple formats (JSON, CSV, TXT)
  • Supports both OpenCV and FFprobe for maximum compatibility
  • Read-only operations - never modifies original files

Installation

  1. Clone the repository:
git clone https://github.com/raqueladaia/video_processor.git
cd video_processor
  1. Install required dependencies:
pip install -r requirements.txt
  1. Run the main program:
python main.py

FFmpeg Installation (Windows)

This project requires FFmpeg to be installed and added to your system PATH. Follow these steps:

Step 0: Download FFmpeg

Download FFmpeg (ffmpeg-git-full.7z) from https://www.gyan.dev/ffmpeg/builds/

Step 1: Find and copy the FFmpeg bin folder path

  1. Extract the downloaded FFmpeg archive to a location like C:\ffmpeg
  2. Inside the extracted folder, find and open the bin folder
    • The full path will look something like C:\ffmpeg\bin
  3. Copy this full path to your clipboard

Step 2: Add the path to the environment variables

  1. Press the Windows key + X and select "System"
  2. Go to "Advanced system settings"
  3. In the System Properties window, click on "Environment Variables"
  4. Under "System variables," find and select the Path variable, then click "Edit"
  5. Click "New" and paste the path you copied in Step 1
  6. Click "OK" on all open windows to save the changes

Step 3: Verify the installation

  1. Open Command Prompt (search for cmd in the Start Menu)
  2. Type the following command and press Enter:
ffmpeg -version
  1. If the installation was successful, you will see the FFmpeg version information displayed

Requirements

  • Python 3.7+
  • FFmpeg (system installation required)
  • OpenCV (cv2)
  • Pandas
  • openpyxl
  • NumPy
  • Pillow

Usage

Main Interface

Run the main script to access both tools:

python main.py

Long Video Chopping

Run directly:

python long_video_chopping/main.py

Features:

  • Choose between single video or folder processing
  • Specify chunk duration in seconds
  • Automatic calculation of number of chunks
  • Progress feedback during processing

Snippet Selection

Run directly:

python snippet_selection/main.py

Features:

  • Excel file parsing for timestamp data
  • Flexible sheet selection: Choose all sheets, specific sheets, or ranges
  • Flexible column name recognition
  • Forward-fills empty video name cells
  • Extracts animal ID from video names (2nd underscore-separated item)
  • Pain/non-pain classification from attention columns
  • Configurable before/after duration (default: -5s/+10s)
  • Fast FFmpeg-based extraction
  • Missing video handling
  • Already processed video detection
  • CSV report generation

Video Metadata Check

Run directly:

python video_metadata_check/main.py

Features:

  • Single or batch analysis: Analyze one video or entire directories
  • Comprehensive metadata extraction:
    • Recording frame rate (from codec)
    • Playback frame rate
    • Actual frame rate (calculated from frame count/duration)
    • Duration and length
    • Number of frames
    • Resolution (width x height)
    • File size
    • Video/audio codecs
    • Bitrate
  • Comparison mode: Compare multiple videos for consistency
    • Select which fields to compare
    • Validate against user-specified criteria
    • Detect anomalies and outliers
  • Multiple output formats:
    • Console output with formatted tables
    • JSON export for programmatic use
    • CSV export for spreadsheets
    • Human-readable TXT reports

Excel File Format

For snippet selection, the Excel file should contain:

Sheet Selection: The program will ask which sheets to process:

  • Process all sheets at once
  • Select specific sheets by number (e.g., 1, 1,3, 1-3, or all)
  • Supports individual selections, comma-separated lists, and ranges

Video Name Matching: The program uses intelligent matching strategies:

  1. Exact match: Video name in Excel exactly matches file name (without extension)
  2. Partial match: Video name in Excel is contained in file name
    • Example: Excel has 2522_2616_bs → matches file 2522_2616_bs_recording_001.mp4
  3. Fuzzy match: Similar names using similarity scoring (for typos/variations)

Supported Video Formats: .mp4, .avi, .mov, .mkv, .wmv, .flv, .webm, .m4v

  • Works with videos that have no audio track

Required columns (any of these names):

  • Video names: video, video_name, file, filename, file_name
    • Empty cells automatically inherit the value from the previous row
    • Animal ID is extracted from the 2nd underscore-separated segment (e.g., 2522_2616_bs → animal_id: 2616)
  • Timestamps: time, timestamp, time_of_interest, start_time, time_awakening_onset

Optional columns:

  • Pain/Non-pain classification: arousal, arousal_type, type, category, attention_to_left_hindpaw, attention_to_left_paw
    • y or Y = pain
    • n or N = nonpain
    • Empty = unclassified
  • Comments: comment, comments, description, notes

Timestamp formats supported:

  • With parentheses: (4:52:13) or (10:23:45)
  • Without parentheses: 0:02:05 or 2:05
  • Seconds: 125.5

Use Cases

Pain vs Non-Pain Awakenings Analysis

The snippet selection module is specifically designed for analyzing pain and non-pain awakenings in animal behavior studies:

  1. Input: Excel file with awakening timestamps and pain/non-pain classifications

    • Supports multiple animals across different videos
    • Automatically extracts animal IDs from video filenames
    • Handles both classified and unclassified awakenings

  2. Processing:

    • Searches recursively through video directories
    • Extracts snippets around each awakening event
    • Uses FFmpeg for fast, lossless extraction
    • Default: 5 seconds before + 10 seconds after each timestamp

  3. Output:

    • Organized snippets with descriptive filenames
    • CSV report for tracking and analysis
    • Ready for behavioral annotation or machine learning

Output Structure

Long Video Chopping

output_directory/
├── video1_name/
│   ├── video1_name_001.mp4
│   ├── video1_name_002.mp4
│   └── ...
└── video2_name/
    ├── video2_name_001.mp4
    └── ...

Snippet Selection

Snippets are named using the format: {animal_id}_{pain|nonpain}_{timestamp}.{ext}

output_directory/
├── 2616_pain_045213.mp4              # Animal 2616, pain awakening at 04:52:13
├── 3007_nonpain_033207.mp4           # Animal 3007, non-pain awakening at 03:32:07
├── 3008_pain_033425.mp4              # Animal 3008, pain awakening at 03:34:25
├── 2616_073607.mp4                   # Animal 2616, unclassified awakening at 07:36:07
├── snippet_processing_report_YYYYMMDD_HHMMSS.csv
└── ...

Video Metadata Check

Reports are generated in multiple formats with timestamped filenames:

output_directory/
├── video_metadata_20231115_143022.json        # Structured JSON report
├── video_metadata_20231115_143022.csv         # Spreadsheet-compatible CSV
└── video_metadata_report_20231115_143022.txt  # Human-readable text report

Example console output:

================================================================================
VIDEO METADATA REPORT
================================================================================
Total videos analyzed: 3
Report generated: 2023-11-15 14:30:22
================================================================================

[Video 1] experiment_video_001.mp4
--------------------------------------------------------------------------------
  Recording Frame Rate........... 30.00 FPS
  Playback Frame Rate............ 30.00 FPS
  Actual Frame Rate.............. 29.97 FPS
  Duration....................... 00:05:30
  Number of Frames............... 9,900
  Resolution..................... 1920x1080
  File Size...................... 245.67 MB

================================================================================
COMPARISON RESULTS
================================================================================

✓ Fields that MATCH across all videos:
  • Frame Rate (FPS): 30.00
  • Resolution: 1920x1080
  • Video Codec: h264

✗ Fields that DIFFER across videos:
  • Duration:
    - 00:05:30: 2 video(s)
    - 00:10:15: 1 video(s)

Project Structure

video_processing/
├── main.py                          # Main entry point
├── requirements.txt                 # Dependencies
├── shared/                          # Common utilities
│   ├── video_utils.py              # Video operations
│   ├── file_utils.py               # File handling
│   └── user_interface.py           # UI functions
├── long_video_chopping/            # Video chopping module
│   ├── main.py                     # Entry point
│   └── video_processor.py          # Core processing
├── snippet_selection/              # Snippet extraction module
│   ├── main.py                     # Entry point
│   ├── excel_parser.py             # Excel file handling
│   ├── video_extractor.py          # Video extraction
│   ├── csv_manager.py              # CSV reporting
│   └── file_manager.py             # File discovery
├── adjust_brightness/              # Brightness adjustment module
│   └── main.py                     # GUI application
├── crop_video/                     # Video cropping module
│   └── main.py                     # GUI application
└── video_metadata_check/           # Video metadata analysis module
    ├── main.py                     # Entry point
    ├── metadata_extractor.py       # Metadata extraction
    ├── metadata_comparator.py      # Video comparison
    └── report_generator.py         # Report generation

Error Handling

The toolkit includes comprehensive error handling:

  • Input validation for all user inputs
  • Video file format verification
  • Missing file detection and reporting
  • Progress feedback and status updates
  • Graceful handling of processing errors

Contributing

This project follows these coding principles:

  • Comprehensive commenting for code understanding
  • Never modify original files
  • Avoid redundant code
  • Reuse existing functions when possible

License

This project is open source. Please refer to the repository for license details.

Support

For issues and feature requests, please use the GitHub repository issue tracker.

About

IN DEVELOPMENT. This repository allows you to modify videos.

Resources

Readme
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages