基于python实现百度语音识别和图灵对话

图例如下

https://github.com/Dongvdong/python_Smartvoice

• 上电后，只要周围声音超过 2000，开始录音5S

• 录音上传百度识别，并返回结果文字输出

• 继续等待，周围声音是否超过2000，没有就等待。

• 点用电脑API语音交互

代码如下

# -*- coding: utf-8 -*-
# 树莓派
from pyaudio import PyAudio, paInt16
import numpy as np
from datetime import datetime
import wave
import time
import requests#导入requests库
import urllib, urllib.request, pycurl
import base64
import json
import os
import sys
from imp import reload

# 调用电脑API生成语音交互
import speech
import win32api
import os
import sys
import time
import win32con

reload(sys)

#sys.setdefaultencoding( "utf-8" )
#一些全局变量
save_count = 0
save_buffer = []
t = 0
sum = 0
time_flag = 0
flag_num = 0
filename = ''
duihua = '1'
def getHtml(url):
 html= requests.get(url)
 # html.encoding = 'utf-8'#防止中文乱码

return html.text
def get_token():
 apiKey = "AxXDYEN27Ks9XHocsGmCEdPm"
 secretKey = "61cd52759f4d704d91c155a22ff7183d"
 auth_url = "https://openapi.baidu.com/oauth/2.0/token?grant_type=client_credentials&client_id=" + apiKey + "&client_secret=" + secretKey;
 res = requests.get(auth_url)
 #res.encoding = 'utf-8'#防止中文乱码
 #print (res.text)
 return json.loads(res.text)['access_token']
def dump_res(buf):#输出百度语音识别的结果
 global duihua
 #print ("字符串类型")
 #print (buf)
 a = eval(buf)
 #print (type(a))
 if a['err_msg']=='success.':
   #print (a['result'][0]）#终于搞定了，在这里可以输出，返回的语句
   duihua = a['result'][0]
   print ("我："+duihua)
def use_cloud(token):#进行合成
 fp = wave.open(filename, 'rb')
 nf = fp.getnframes()
 f_len = nf * 2
 audio_data = fp.readframes(nf)
 cuid = "9120612" #产品id
 srv_url = 'http://vop.baidu.com/server_api' + '?cuid=' + cuid + '&token=' + token
 http_header = [
   'Content-Type: audio/pcm; rate=8000',
   'Content-Length: ％d' ％ f_len
 ]
 c = pycurl.Curl()
 c.setopt(pycurl.URL, str(srv_url)) #curl doesn't support unicode
 #c.setopt(c.RETURNTRANSFER, 1)
 c.setopt(c.HTTPHEADER, http_header)  #must be list, not dict
 c.setopt(c.POST, 1)
 c.setopt(c.CONNECTTIMEOUT, 30)
 c.setopt(c.TIMEOUT, 30)
 c.setopt(c.WRITEFUNCTION, dump_res)
 c.setopt(c.POSTFIELDS, audio_data)
 c.setopt(c.POSTFIELDSIZE, f_len)
 c.perform() #pycurl.perform() has no return val
# 将data中的数据保存到名为filename的WAV文件中
def save_wave_file(filename, data):
 wf = wave.open(filename, 'wb')
 wf.setnchannels(1)
 wf.setsampwidth(2)
 wf.setframerate(SAMPLING_RATE)
 wf.writeframes(b"".join(data))
 wf.close()
NUM_SAMPLES = 2000    # pyAudio内部缓存的块的大小
SAMPLING_RATE = 8000  # 取样频率
LEVEL = 1500      # 声音保存的阈值
COUNT_NUM = 20     # NUM_SAMPLES个取样之内出现COUNT_NUM个大于LEVEL的取样则记录声音
SAVE_LENGTH = 8     # 声音记录的最小长度：SAVE_LENGTH * NUM_SAMPLES 个取样
exception_on_overflow=False
# 开启声音输入ｐｙａｕｄｉｏ对象
pa = PyAudio()
stream = pa.open(format=paInt16, channels=1, rate=SAMPLING_RATE, input=True,
       frames_per_buffer=NUM_SAMPLES)
token = get_token()#获取ｔｏｋｅｎ
key = '35ff2856b55e4a7f9eeb86e3437e23fe'
api = 'http://www.tuling123.com/openapi/api?key=' + key + '&info='
while(True):
 # 读入NUM_SAMPLES个取样
 string_audio_data = stream.read(NUM_SAMPLES,False);
 # 将读入的数据转换为数组
 audio_data = np.fromstring(string_audio_data, dtype=np.short)
 # 计算大于LEVEL的取样的个数
 large_sample_count = np.sum( audio_data > LEVEL )
 temp = np.max(audio_data)
 if temp > 2000 and t == 0:
   t = 1#开启录音
   print ("---------主人我在听你说！（5S）----------")
   begin = time.time()
   # print (temp)
 if t:
   #print (np.max(audio_data))
   if np.max(audio_data)<1000:
     sum += 1
     # print (sum)
   end = time.time()
   if end-begin>5:
     time_flag = 1
     # print ("五秒到了，准备结束")
   # 如果个数大于COUNT_NUM，则至少保存SAVE_LENGTH个块
   if large_sample_count > COUNT_NUM:
     save_count = SAVE_LENGTH
   else:
     save_count -= 1
   if save_count < 0:
     save_count = 0
   if save_count > 0:
     # 将要保存的数据存放到save_buffer中
     save_buffer.append(string_audio_data )
   else:
     # 将save_buffer中的数据写入WAV文件，WAV文件的文件名是保存的时刻
     #if time_flag:
     if len(save_buffer) > 0 or time_flag:
       #filename = datetime.now().strftime("％Y-％m-％d_％H_％M_％S") + ".wav"#原本是用时间做名字
       filename = str(flag_num)+".wav"
       flag_num += 1
       save_wave_file(filename, save_buffer)
       save_buffer = []
       t = 0
       sum =0
       time_flag = 0
      # print (filename, "保存成功正在进行语音识别")
       use_cloud(token)
      #  print (duihua)
       info = duihua
       duihua = ""

request = api + str(info)
       response = getHtml(request)
      # print ( "-----1-----")
       dic_json = json.loads(response)

a = dic_json['text']

unicodestring = a
       # 将Unicode转化为普通Python字符串："encode"
       utf8string = unicodestring.encode("utf-8")

print ("科塔娜："+str(a))

# 电脑说话
       speech.say(str(a))

url = "http://tsn.baidu.com/text2audio?tex="+dic_json['text']+"&lan=zh&per=0&pit=1&spd=7&cuid=7519663&ctp=1&tok=25.41bf315625c68b3e947c49b90788532d.315360000.1798261651.282335-9120612"
       os.system('mpg123 "％s"'％(url))

来源：https://www.cnblogs.com/kekeoutlook/p/9386174.html