视频文字提取（方法二）

结合DeepSeek给出的第二种提取文字的方法代码如下：

from moviepy.editor import VideoFileClip

# 视频文件路径
video_path = '视频文件路径'

# 加载视频文件
video = VideoFileClip(video_path)

# 从视频中提取音频部分
audio = video.audio

# 保存音频为临时文件
audio_path = 'temp_audio.wav'
audio.write_audiofile(audio_path)

from pydub import AudioSegment

# 载入音频文件
audio = AudioSegment.from_wav(audio_path)

# 将音频转换为单声道并设置适当的采样率
audio = audio.set_channels(1)
audio = audio.set_frame_rate(16000)

# 存储转换后的音频文件
processed_audio_path = '保存音频文件路径/temp_audio.wav'
audio.export(processed_audio_path, format="wav")

import speech_recognition as sr

# 初始化识别器
recognizer = sr.Recognizer()

# 从转换后的音频文件中加载数据
with sr.AudioFile(processed_audio_path) as source:
    audio_data = recognizer.record(source)

# 识别音频中的语音内容
try:
    text = recognizer.recognize_google(audio_data, language='zh-CN')  # 假设音频语言为中文
    print(text)

    # 将识别的文字保存到文本文件
    txt_file_path = '保存文字结果路径/result.txt'  # 设定保存文本的路径
    with open(txt_file_path, 'w', encoding='utf-8') as f:
        f.write(text)
    print(f"识别结果已保存到 {txt_file_path}")

except sr.UnknownValueError:
    print("Google Speech Recognition could not understand audio")
except sr.RequestError:
    print("Could not request results from Google Speech Recognition service")

# 清理临时文件
import os

os.remove(audio_path)
os.remove(processed_audio_path)