auto-video-trimming-export

Overview

This project aims to automate the trimming and exporting of video files.

Notes

Due to the incomplete video analysis process, this program is not generic.
In the future, I’d like to support analysis processing based on certain rules.
At this time, please use the method of specifying time information in setting.toml to execute the program.

Approaches to Automation and Processing Flow

MoviePy and QuickTime are used for trimming and exporting.

The key to automation is "creating the time information to be specified for trimming," which requires analysis of the video file (frame information).

OpenCV is used to extract frame information, and analysis processing is performed in units of seconds according to the specifications in the TOML file.

The analysis process obtains timestamps corresponding to frames that match the conditions and saves them as trimming data in a list.

Features

Trimming can be performed on video files according to the following specifications.

• Detects frames in a video file using a specified method, cuts off the detected portions, and saves them as a separate file.
  • Average of brightness data
  • Character recognition by OCR(pytesseract)(*1)
  • Use Image comparison by histogram correlation coefficient(*1)
• Trimming data output (csv file)

(*1) Can be used together

Usage

1. Get project

% git clone https://github.com/7rikazhexde/auto-video-trimming-export.git 

2. Trimming Process Settings

This project uses the information defined in setting.toml to read, log, analyze, create timestamps, and process the export of video files.

• [log]

  Define settings for outputting execution results to the log and console.

• [media]

  Defines the video file path to be analyzed and processed. This program targets MOV and MP4 operation. Others are not guaranteed.

• [[parse_info]]

  parse_info[0] defines whether the analysis process can be performed or not.

  parse_info[1] and thereafter define variables to be used in the analysis process.

  For example, if you want to analyze three different times, define elements 1 through 3.

• [trim_info]

  Trimming processing time
  Set the time information in the time_list key (list) in [trim_info] of setting.toml.

  Trimming processing time setting based on video analysis
  Set an empty time information in the time_list key (list) in [trim_info] of setting.toml.

Trimming and exporting process

For Mac, use QuickTime Player from AppleScript.

For Windows and linux, use MoviePy.

Each process is automatically detected and processed based on platform information.

MoviePy can be used for Mac, but QuickTime Player is used because QuickTime Player is faster for export processing.

If you want to use MoviePy on a Mac, change self.__pf to something other than 'Darwin'.

Notes

• Output file resolution is 640X360.

• Codec specification is as follows

  Program	codec	codec audio
  AppleScript	H264	AAC
  MoviePy	libx264	AAC

• Except for the resolution setting (480p), the default settings will be used.

• MoviePy more details

3. Setup of virtual environment Run the poetry command.

% poetry install

• If the package DL fails after installation, there may be a problem with the development environment.
• See Switching between environments.
• Please run poetry env info to check your development environment.
• If your python version is not 3.10 or higher, please run poetry env use python3.10 to recreate your development environment.

Or create a virtual environment with venv, pyenv, etc. and run the following command.

% pip install -r requirements.txt

3. Execute the program in a virtual environment

% cd app
% poetry shell
% python main.py

4. When executing a single trimming process

• To use AppleScript (Mac only)

% cd app
% python exe_applescript.py

• To use MoviePy

% cd app
% python moviepy_trim.py

Other Information

If you are using the lossless-cut application for the trimming process, the timestamp lists output by this program (app/media/timestamp_data.csv) can be used with the import function.

Example

0.0,11.011000000000001,seg_1
62.99626666666667,103.00290000000001,seg_2