视频转文字-视频语义分析

Project ：视频转文本
Introduction：通过使用FFmpeg-Python库,将视频的音频抽取出来存放至本地文件夹，再调用
百度语音识别REST Api，将音频转化文字，该Api可识别英语和普通话
Attention ：上传的视频不能超过60s.同时需要主机联网才可调用云Api
Quickstart :
1.Download Anaconda and Install it
https://www.anaconda.com/distribution/
2.Open the Terminal,and input this command:
conda env create -f bat_video.yaml
3. python main.py

import os
import ffmpeg
from aip import AipSpeech


# 百度语音识别REST-Api 
APP_ID = '176xxxx'
API_KEY = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'
SECRET_KEY = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'
client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)

# 输入视频路径
video_path = os.path.abspath('./Raw_Video')
text_path =  os.path.abspath('./Output_Text')
radio_path = os.path.abspath('./Raw_Radio')
files= os.listdir(video_path)


# 主函数
def main():
	for file in files: 
		filename = os.path.splitext(str(file))[0]
		# ffmpeg视频抽取音频
		in_stream = ffmpeg.input(video_path+'/'+str(file))
		(in_stream   
		    .output(radio_path+'/'+filename+'.pcm',format='s16le', acodec='pcm_s16le', ac=1, ar='16k')
		    .run()
		)
		# 识别本地文件
		test=client.asr(get_file_content(radio_path+'/'+filename+'.pcm'), 'pcm', 16000, {
    
    
		    'dev_pid': 1537,
		})
        # 内容写入文本并保存
		txt_name = text_path+'/'+filename+'.txt'
		f = open(txt_name, 'w')
		f.write(str(test['result']))
		f.close()


# 读取文件
def get_file_content(filePath):
    with open(filePath, 'rb') as fp:
        return fp.read()


if __name__=='__main__':
	main()