本文目录导读：
使用 Whisper + FFmpeg（推荐）
这个方案最成熟,支持多种语言：
# 方法1: Python脚本
import whisper
import subprocess
import os
def add_subtitles(video_path, output_path):
# 1. 语音识别
model = whisper.load_model("base") # 可选: tiny, base, small, medium, large
result = model.transcribe(video_path)
# 2. 保存为SRT字幕文件
srt_path = video_path.replace('.mp4', '.srt')
with open(srt_path, 'w', encoding='utf-8') as f:
for i, segment in enumerate(result['segments'], start=1):
start = format_time(segment['start'])
end = format_time(segment['end'])
f.write(f"{i}\n{start} --> {end}\n{segment['text']}\n\n")
# 3. 用FFmpeg嵌入字幕
cmd = f'ffmpeg -i "{video_path}" -vf "subtitles={srt_path}" "{output_path}"'
subprocess.run(cmd, shell=True)
def format_time(seconds):
h = int(seconds // 3600)
m = int((seconds % 3600) // 60)
s = int(seconds % 60)
ms = int((seconds - int(seconds)) * 1000)
return f"{h:02d}:{m:02d}:{s:02d},{ms:03d}"
# 使用示例
add_subtitles("input.mp4", "output.mp4")
使用命令行工具
#!/bin/bash # Linux/macOS 脚本 # 安装依赖 pip install openai-whisper ffmpeg-python # 使用whisper命令行 whisper input.mp4 --model base --output_format srt # 嵌入字幕到视频 ffmpeg -i input.mp4 -vf "subtitles=input.srt" output_with_sub.mp4
使用阿里云/腾讯云API
# 使用阿里云语音识别API
import json
import requests
from aliyunsdkcore.client import AcsClient
def add_subtitles_with_cloud(video_path):
# 1. 上传音频到OSS
# 2. 调用语音识别API
# 3. 获取字幕结果
# 4. 合成到视频
pass
使用MoviePy（适合简单场景）
from moviepy.editor import VideoFileClip, TextClip, CompositeVideoClip
def add_simple_subtitles(video_path, subtitles_list):
"""
subtitles_list: [(start_time, end_time, text), ...]
"""
video = VideoFileClip(video_path)
txt_clips = []
for start, end, text in subtitles_list:
txt_clip = TextClip(text, fontsize=24, color='white',
font='simhei.ttf', stroke_color='black')
txt_clip = txt_clip.set_position(('center', 'bottom')).set_duration(end-start).set_start(start)
txt_clips.append(txt_clip)
final = CompositeVideoClip([video] + txt_clips)
final.write_videofile("output.mp4")
安装依赖
# 安装FFmpeg (Mac) brew install ffmpeg # 安装FFmpeg (Ubuntu) sudo apt-get install ffmpeg # 安装Python包 pip install openai-whisper ffmpeg-python moviepy
完整自动化脚本
#!/usr/bin/env python3
import os
import sys
import subprocess
import whisper
from pathlib import Path
class AutoSubtitle:
def __init__(self, model_size="base"):
self.model = whisper.load_model(model_size)
def process_video(self, video_path, language=None):
# 1. 转音频
audio_path = video_path.replace('.mp4', '.wav')
subprocess.run([
'ffmpeg', '-i', video_path,
'-vn', '-acodec', 'pcm_s16le',
'-ar', '16000', '-ac', '1', audio_path
])
# 2. 语音识别
options = {}
if language:
options['language'] = language
result = self.model.transcribe(audio_path, **options)
# 3. 生成SRT
srt_path = video_path.replace('.mp4', '.srt')
self._save_srt(result, srt_path)
# 4. 嵌入字幕
output_path = video_path.replace('.mp4', '_subtitled.mp4')
subprocess.run([
'ffmpeg', '-i', video_path,
'-vf', f"subtitles={srt_path}",
'-c:a', 'copy', output_path
])
# 清理临时文件
os.remove(audio_path)
return output_path
def _save_srt(self, result, srt_path):
with open(srt_path, 'w', encoding='utf-8') as f:
for i, segment in enumerate(result['segments'], 1):
start = self._format_time(segment['start'])
end = self._format_time(segment['end'])
f.write(f"{i}\n{start} --> {end}\n{segment['text'].strip()}\n\n")
def _format_time(self, seconds):
h = int(seconds // 3600)
m = int((seconds % 3600) // 60)
s = int(seconds % 60)
ms = int((seconds - int(seconds)) * 1000)
return f"{h:02d}:{m:02d}:{s:02d},{ms:03d}"
# 使用
if __name__ == "__main__":
subtitle = AutoSubtitle("base")
output = subtitle.process_video("input.mp4", language="zh")
print(f"字幕已添加到: {output}")
批量处理脚本
#!/bin/bash
# 批量处理所有MP4文件
for video in *.mp4; do
echo "处理: $video"
python3 -c "
import auto_subtitle
s = auto_subtitle.AutoSubtitle()
s.process_video('$video')
"
done
注意事项
- 模型选择：
tiny最快但准确度较低，large最准确但需要更多资源 - GPU加速：如果有NVIDIA GPU，加上
--device cuda - 语言指定：中文视频加
--language zh - 字幕样式：可以通过FFmpeg的
subtitles滤镜调整样式
这个方案最适合中文视频,准确度能达到90%以上。
