python音色模型 python声音模块

安装

pip install -i https://pypi.douban.com/simple/ pyttsx3

简单使用：

import pyttsx3

pyttsx3.speak('泸州彭于晏仅次于彭于晏，哈哈哈哈')

使用引擎进行播报，并保存为本地语音文件：

import pyttsx3

engine = pyttsx3.init()  # object creation

""" RATE"""
rate = engine.getProperty('rate')  # 获取当前语速的细节
print(rate)  # 打印当前语音速率
engine.setProperty('rate', 125)  # 设置新的语音速率

"""VOLUME"""
volume = engine.getProperty('volume')  # 了解当前音量级别(min=0, max=1)
print(volume)  # 打印电流量级
engine.setProperty('volume', 1.0)  # 将卷级别设置在0到1之间

"""VOICE"""
voices = engine.getProperty('voices')  # 获取当前语音的细节
engine.setProperty('voice', voices[0].id)  # 改变指数，改变声音。o为男性
# engine.setProperty('voice', voices[1].id)  # 改变指数，改变声音。1女

engine.say("Hello World!")
engine.say('My current speaking rate is ' + str(rate))
engine.runAndWait()
engine.stop()

"""Saving Voice to a file"""
# 在linux上，确保安装了' speak'和'ffmpeg'
engine.save_to_file('Hello World', 'test.mp3')  # 将其存储为语音文件
engine.runAndWait()

在网上看到一篇语音助手文章，所以制作一款语音播报助手：

import datetime
import json
import cv2
import pyttsx3
from lxml import etree
import requests


def get_weather():
    url = 'http://www.weather.com.cn/weather/101290101.shtml'
    response = requests.get(url)
    response.encoding = 'utf-8'
    response = response.text  # 获取页面
    html = etree.HTML(response)
    day_weather = '天气状况：' + html.xpath('//*[@id="7d"]/ul/li[1]/p[1]/text()')[0] + '\n'  # 获取天气，白天的天气
    high = html.xpath('//*[@id="7d"]/ul/li[1]/p[2]/span/text()')
    low = html.xpath('//*[@id="7d"]/ul/li[1]/p[2]/i/text()')  # 获取对应的两个温度
    # 因为页面在晚上会有小变化，所以使用条件语句，来排除因变化引起的bug
    if high == []:
        day_temperature = '室外温度：' + low[0] + '\n'
    else:
        day_temperature = '室外温度：' + low[0].replace('℃', '') + '~' + high[0] + '℃\n'  # 获取温度
    # 获取两个风向
    wind_1 = html.xpath('//*[@id="7d"]/ul/li[1]/p[3]/em/span[1]/@title')
    wind_2 = html.xpath('//*[@id="7d"]/ul/li[1]/p[3]/em/span[2]/@title')
    # 因为有时候，会出现两个风向是同一个风向的情况，所以使用条件语句排除
    if wind_2 == []:
        wind = wind_1[0] + '\n'
    elif wind_1[0] == wind_2[0]:
        wind = wind_1[0] + '\n'
    else:
        wind = wind_1[0] + '转' + wind_2[0] + '\n'
    # 因为风级有时候会出现“<"，语音的时候会认为是爱心符号，所以使用替换，改为文字”低于“
    wind_3 = html.xpath('//*[@id="7d"]/ul/li[1]/p[3]/i/text()')[0].replace('<', '低于').replace('>', '高于')
    day_wind = '风向情况：' + wind + wind_3 + '\n'  # 获取风向及风级
    return day_weather, day_temperature, day_wind


def get_time():
    a = datetime.datetime.now()  # 实施时间
    y = str(a.year)
    m = str(a.month)
    d = str(a.day)  # 转换为字符串，便于打印
    time = y + '年' + m + '月' + d + '日' + '\n'
    b = datetime.datetime(2023, 4,27)  # 自己设置的专升本时间
    count_down = (b - a).days  # 专升本倒计时
    return time, count_down


def get_content():
    url = 'http://open.iciba.com/dsapi/'  # 网上找的API
    response = requests.get(url=url)
    json_s = json.loads(response.text)
    jitang = json_s.get("content") + '\n'  # 每日鸡汤
    translation = json_s.get("note") + '\n'  # 中文翻译
    image_url = json_s.get("fenxiang_img")  # 图片链接
    return jitang, translation, image_url


def main():
    time, count_down = get_time()
    day_weather, day_temperature, day_wind = get_weather()
    jitang, translation, image_url = get_content()
    count_down = '距离专升本还有{}天，你准备好了吗？'.format(count_down) + '\n'
    a = '下面为您播报今日天气状况\n'
    b = '每日一句\n'
    time = '今天是' + time
    weather = day_weather + day_temperature + day_wind
    content = jitang + translation
    text = time + count_down + a + weather + b + content  # 语音内容
    voice = pyttsx3.init()  # 初始化
    # rate = voice.getProperty('rate')
    voice.setProperty('rate', 150)  # 语速，范围在0-200之间
    voice.setProperty('volume', 1.0)  # 范围在0.0-1.0之间
    voice.say(text)  # 语音内容
    voice.runAndWait()
    cap = cv2.VideoCapture(image_url)  # 展示图片
    if (cap.isOpened()):
        ret, img = cap.read()
        my_image = cv2.resize(img, dsize=None, fx=0.5, fy=0.5)
        cv2.imshow("You will succeed in the end", my_image)
        cv2.waitKey()
    print(time, weather, content)


if __name__ == '__main__':
    main()