'''
Created on 2012-7-2
@author: Administrator
'''
import wx
import numpy as np
import wx.lib.plot as wxPyPlot # 导入绘图模块,并命名为wxPyPlot
import wave
# 需要把数据封装进入MyDataObject中
def MyDataObject():
# 50 个点的sin函数,用蓝色圆点表示
data1 = 2.*np.pi*np.arange(100)/100.
data1.shape = (50, 2)
data1[:, 1] = np.sin(data1[:, 0])
print ("debug:", data1.shape)
markers = wxPyPlot.PolyMarker(data1, legend='Green Markers', colour='blue', marker='circle', size=1)
# 50个点的cos函数,用红色表示
data2 = 2.*np.pi*np.arange(100)/100.
data2.shape = (50, 2)
print ("debug: data2", len(data2))
data2[:, 1] = np.cos(data2[:, 0])
lines = wxPyPlot.PolySpline(data2, legend='Red Line', colour='red')
GraphTitle = "Plot Data(Sin and Cos)"
return wxPyPlot.PlotGraphics([markers, lines], GraphTitle, "X Axis", "Y Axis")
# 解析wav数据
def MyWavData(wav_filename=""):
print('working')
# 打开wav文档
file = wave.open("mic4.wav", "r")
# 读取格式信息
# (nchannels, sampwidth,framerate, nframes, comptype, compname)
params = file.getparams()
nchannels, sampwidth, framerate, nframes = params[:4]
print (nchannels, sampwidth, framerate, nframes)
# 读取波形数据
str_data = file.readframes(nframes)
# 文件使用完毕,关闭文件
file.close()
# 将波形数据装换成数组
wave_data = np.fromstring(str_data, dtype=np.short)
wave_data.shape = (-1, 2)
wave_data = wave_data.T # 矩阵转置
time = np.arange(0, nframes) * (1.0 / framerate)
# print ("debug: time:", len(time))
# print ("debug: wave_data:", len(wave_data[0][0:len(time)]))
# print ("debug: time:", time)
# print ("debug: wave:", wave_data)
time_and_wav = np.asarray([time, wave_data[0][0:len(time)]]).T
print ("debug: len of time and wav: ", len(time_and_wav))
print ("debug: time and wav: ", time_and_wav.shape)
lines = wxPyPlot.PolySpline(time_and_wav, legend='Blue Line', colour='blue')
GraphTitle = "the freq of wav file"
return wxPyPlot.PlotGraphics([lines, ], GraphTitle, "time/s", "fre/Hz")
class TestFrame1(wx.Frame):
def __init__(self, parent=None, id=wx.ID_ANY, title="Using wxPyPlot"):
wx.Frame.__init__(self, parent, id, title)
# # 创建菜单栏
# self.mainmenu = wx.MenuBar()
# # 创建菜单
# menu = wx.Menu()
# menu.Append(100, 'Draw1', 'Draw plots1')
# self.Bind(wx.EVT_MENU, self.OnPlotDraw1, id=100)
# menu.Append(200, 'Draw_wav', 'Draw wav')
# self.Bind(wx.EVT_MENU, self.OnPlotDraw_wav, id=200)
# # 添加菜单到菜单栏
# self.mainmenu.Append(menu, '&Plot')
# # 设置菜单Bar
# self.SetMenuBar(self.mainmenu)
# # 创建状态栏,显示信息
self.CreateStatusBar(2)
self.pc = wxPyPlot.PlotCanvas(self) # 此处导入绘图面板
self.pc.Draw(MyWavData())
def OnPlotDraw1(self, event): # 绘图函数
self.pc.Draw(MyDataObject())
def OnPlotDraw_wav(self, event):
self.pc.Draw(MyWavData())
class Example(wx.Frame):
def __init__(self, parent, title):
super(Example,self).__init__(parent, title=title, size=(1000,600))
self.InitUI()
self.Centre()
self.Show()
def InitUI(self):
panel = wx.Panel(self)
sizer = wx.GridBagSizer(12, 12)
# text = wx.StaticText(panel, label='Rename to')
# sizer.Add(text, pos=(0, 0), flag=wx.TOP|wx.LEFT|wx.BOTTOM, border=5)
#
# tc = wx.TextCtrl(panel)
# sizer.Add(tc, pos=(0,1), span=(1,5), flag=wx.EXPAND, border=5)
# 这里是原始语音
tf = wx.Button(panel, label="原始语音")
sizer.Add(tf, pos=(1, 1), span=(2, 12), flag = wx.EXPAND|wx.ALL)
# 这里是增强
up = wx.Button(panel, label="增强语音")
sizer.Add(up, pos=(3, 1), span=(4, 12), flag = wx.EXPAND|wx.ALL)
# 这里是识别文字
cog = wx.Button(panel, label="识别文字")
sizer.Add(cog, pos=(9, 1), span=(3, 12), flag = wx.EXPAND|wx.ALL)
# 其他布局
buttonLuyin = wx.Button(panel,label='录音',size=(90,28))
buttonClose = wx.Button(panel,label='播放',size=(90,28))
buttonRecog = wx.Button(panel,label='识别',size=(90,28))
sizer.Add(buttonLuyin, pos=(12, 1))
sizer.Add(buttonClose, pos=(12, 2))
sizer.Add(buttonRecog, pos=(12, 3), flag=wx.RIGHT|wx.BOTTOM, border=5)
sizer.AddGrowableRow(2)
sizer.AddGrowableCol(1)
panel.SetSizerAndFit(sizer)
if __name__ == '__main__':
app = wx.App()
Example(None, title="语音识别展示")
app.MainLoop()//test
#!bin/bash/python
# -*- coding=utf-8 -*-
from __future__ import division
import wx
import wx.media
import os.path, os, gc, sys, pyaudio
import wave, numpy, codecs, time
import pylab as pl
import pyaudio
#import win32gui
from matplotlib.ticker import MultipleLocator, FormatStrFormatter
import sys
reload(sys)
sys.setdefaultencoding("utf-8")
global wavCount
wavCount=1
input_filename = "input.wav" # 麦克风采集的语音输入
input_filepath = "./" # 输入文件的path
in_path = input_filepath + input_filename
global_frames = []
xmajorLocator = MultipleLocator(0.5) #将x主刻度标签设置为0.5的倍数
xmajorFormatter = FormatStrFormatter('%1.1f') #设置x轴标签文本的格式
xminorLocator = MultipleLocator(0.05) #将x轴次刻度标签设置为0.05的倍数
class MyFrame(wx.Frame):
#初始化全部界面
def __init__(self, wavlist, recList):
self.wavTime=0.000
self.wavLength=0.000
self.labelcount=0
self.choosenItem=0
self.labelPosition=0
self.cutCount=1
self.wavList=[]
self.recList=[]
self.labellist=[]
self.labelPos=[]
self.tempwavlist=[]
self.wavList = wavlist
self.recList = recList
# self.wavfilepath = wavlist[0][1]
# self.recTextStr = recList[0][1]
self.wavfilepath = wavlist[0]
self.recTextStr = recList[0]
print ("wav:", self.wavfilepath)
print ("rec:", self.recTextStr)
wx.Frame.__init__(self,None,size=(1920,1080))
self.sp=wx.SplitterWindow(self,style=wx.SP_3D)
self.mainPanel=wx.Panel(self.sp)
self.listPanel=wx.Panel(self.sp)
self.sp2=wx.SplitterWindow(self.mainPanel)
self.annoPanel=wx.Panel(self.sp2)
self.subPanel=wx.Panel(self.sp2)
self.box=wx.BoxSizer(wx.VERTICAL)
self.box.Add(self.sp2,1,wx.EXPAND)
self.sp3=wx.SplitterWindow(self.subPanel)
self.cutPanel=wx.Panel(self.sp3)
self.statisticPanel=wx.Panel(self.sp3)
self.box2=wx.BoxSizer(wx.VERTICAL)
self.box2.Add(self.sp3,1,wx.EXPAND)
self.sp4=wx.SplitterWindow(self.statisticPanel)
self.wavPanel=wx.Panel(self.sp4)
# upgrade
# self.upwavPanel=wx.Panel(self.sp4)
self.infoPanel=wx.Panel(self.sp4)
self.box3=wx.BoxSizer(wx.VERTICAL)
self.box3.Add(self.sp4,1,wx.EXPAND)
self.sp5=wx.SplitterWindow(self.infoPanel)
self.labelPanel=wx.Panel(self.sp5)
self.aaaPanel=wx.Panel(self.sp5)
self.box4=wx.BoxSizer(wx.VERTICAL)
self.box4.Add(self.sp5,1,wx.EXPAND)
self.sp6=wx.SplitterWindow(self.aaaPanel)
self.bbbPanel=wx.Panel(self.sp6)
self.cccPanel=wx.Panel(self.sp6)
self.box5=wx.BoxSizer(wx.VERTICAL)
self.box5.Add(self.sp6,1,wx.EXPAND)
self.mainPanel.SetSizer(self.box)
self.subPanel.SetSizer(self.box2)
self.statisticPanel.SetSizer(self.box3)
self.infoPanel.SetSizer(self.box4)
self.aaaPanel.SetSizer(self.box5)
self.cutPanel.SetBackgroundColour("white")
self.bbbPanel.SetBackgroundColour("white")
self.cccPanel.SetBackgroundColour("white")
print ("debug: wavfilepath:", self.wavfilepath)
self.drawBackJpg(str(self.wavfilepath))
self.CreateListUI(wavlist)
self.CreateMainUI(recList)
self.CreateWavUI()
self.CreateLabelListUI()
self.sp.SplitVertically(self.listPanel,self.mainPanel,150)
self.sp2.SplitHorizontally(self.annoPanel,self.subPanel,300)
self.sp3.SplitHorizontally(self.cutPanel,self.statisticPanel,110)
self.sp4.SplitHorizontally(self.wavPanel,self.infoPanel,280)
self.sp5.SplitHorizontally(self.labelPanel,self.aaaPanel,100)
self.sp6.SplitVertically(self.bbbPanel,self.cccPanel,820)
self.Bind(wx.EVT_LIST_ITEM_SELECTED,self.OnListBox,self.list)
self.wavPanel.Bind(wx.EVT_PAINT, self.OnPaint)
self.wavPanel.Bind(wx.EVT_ERASE_BACKGROUND,self.OnErasePass)
self.wavPanel.Bind(wx.EVT_MOTION,self.OnMove)
self.wavPanel.Bind(wx.EVT_LEFT_DOWN,self.OnLeftDown)
self.wavPanel.Bind(wx.EVT_LEFT_UP,self.OnDrawRec)
# self.upwavPanel.Bind(wx.EVT_PAINT, self.OnPaint)
# self.upwavPanel.Bind(wx.EVT_ERASE_BACKGROUND,self.OnErasePass)
# self.upwavPanel.Bind(wx.EVT_MOTION,self.OnMove)
# self.upwavPanel.Bind(wx.EVT_LEFT_DOWN,self.OnLeftDown)
# self.upwavPanel.Bind(wx.EVT_LEFT_UP,self.OnDrawRec)
self.labelPanel.Bind(wx.EVT_MOTION,self.OnLabelMove)
self.labelPanel.Bind(wx.EVT_LEFT_UP,self.OnLabelDrawLine)
self.labelPanel.Bind(wx.EVT_PAINT,self.OnLabelPaint)
self.labelPanel.Bind(wx.EVT_ERASE_BACKGROUND,self.OnErasePass)
#左侧列表的点击事件
def OnListBox(self,event):
# self.sp.Unsplit()
if self.timer:
self.timer.Stop()
# del self.timer
self.cutCount=1
#获取点击对象
index = event.GetIndex()
self.choosenItem = index
data = self.list.GetItem(index,0)
tuple_temp = self.wavList[int(data.GetText())-1]
# self.wavfilepath = tuple_temp[1]
# self.recTextStr = self.recList[int(data.GetText())-1][1]
self.wavfilepath = self.wavfilepath
self.recTextStr = self.recTextStr
print ("wav:", self.wavfilepath)
print ("rec:", self.recTextStr)
self.OnFreshUI()
#刷新界面
def OnFreshUI(self):
self.InitMedia()
self.rec_tc.SetValue(self.recTextStr)
self.label_tc.SetValue("")
self.anno_tc.SetValue("")
self.wavpath_ST.SetLabel(u"当前操作文件 : "+ self.wavfilepath)
self.drawBackJpg(str(self.wavfilepath))
self.timer = wx.Timer(self)
self.Bind(wx.EVT_TIMER,self.OnTimer)
self.timer.Start(100)
self.cutPanel.Refresh()
self.annoPanel.Refresh()
self.mainPanel.Refresh()
#绘制音频波形的paint事件的回调函数
def OnPaint(self,event):
brush=wx.Brush("white")
img = wx.Bitmap("mic4_output.png")
dc=wx.BufferedPaintDC(self.wavPanel)
dc.SetBackground(brush)
# dc.Clear()
dc.DrawBitmap(img,-172,0)
#据说可以解决屏闪问题
def OnErasePass(self,event):
pass
#--------------------------------------------------
#列表界面的UI创建
def CreateListUI(self,wavlist):
self.ListBox=wx.BoxSizer(wx.HORIZONTAL)
self.list = wx.ListCtrl(self.listPanel, -1, style=wx.LC_REPORT)
self.list.InsertColumn(0, u'序号', width=50)
self.list.InsertColumn(1, u'完成情况', width=150)
for i in wavlist:
index = self.list.InsertStringItem(sys.maxint, str(i[0]))
self.list.SetStringItem(index, 1, u"未完成")
self.ListBox.Add(self.list,1,wx.EXPAND)
self.listPanel.SetSizerAndFit(self.ListBox)
#--------------------------------------------------
#标注界面的UI创建
#wx.media的load初始化
def InitMedia(self):
try:
self.mc = wx.media.MediaCtrl(self.mainPanel, style=wx.SIMPLE_BORDER)
except NotImplementedError:
self.Destroy()
raise
if not self.mc.Load(self.wavfilepath):
wx.MessageBox("Unable to load: Unsupported format?","ERROR", wx.ICON_ERROR | wx.OK)
else:
print "load success"
print self.mc.Length()
self.slider.SetRange(0, self.mc.Length())
#self.mc.Pause=0
#初始完成后各按键才可使用
def OnMediaLoaded(self, evt):
self.buttonPlay.Enable()
#创建标注界面(播放器,文本框)
def CreateMainUI(self,recList):
font=wx.SystemSettings_GetFont(wx.SYS_SYSTEM_FONT)
font.SetPointSize(12)
font20=wx.SystemSettings_GetFont(wx.SYS_SYSTEM_FONT)
font20.SetPointSize(16)
self.annoPanel.SetBackgroundColour("white")
self.MainSizer=wx.GridBagSizer(0,0)
self.num_st=wx.StaticText(self.annoPanel,label="")
self.num_st.SetFont(font)
self.MainSizer.Add(self.num_st,pos=(0,0),span=(1,2),flag=wx.ALL,border=10)
self.num1=wx.StaticText(self.annoPanel,label="",style=wx.ALIGN_CENTER)
self.num1.SetFont(font)
self.MainSizer.Add(self.num1,pos=(2,0),span=(1,2),flag=wx.ALL,border=10)
self.wav_st=wx.StaticText(self.annoPanel,label=u"音频")
self.wav_st.SetFont(font)
self.MainSizer.Add(self.wav_st,pos=(0,2),span=(1,5),flag=wx.EXPAND|wx.ALL,border=10)
self.MainSizer.AddGrowableCol(6,1)
self.buttonPlay=wx.Button(self.annoPanel,wx.ALL,label=u"&▷\tF1")
self.buttonPause=wx.Button(self.annoPanel,label=u"&‖\tEsc")
self.buttonStop=wx.Button(self.annoPanel,label=u"&■\tCtrl-S")
self.Bind(wx.EVT_BUTTON,self.playWav,self.buttonPlay)
#accelTb1 = wx.AcceleratorTable([wx.ACCEL_CTRL,ord('S'),self.buttonPlay.GetId()])
#self.SetAcceleratorTable(accelTb1)
self.Bind(wx.EVT_BUTTON,self.pauseWav,self.buttonPause)
self.Bind(wx.EVT_BUTTON,self.stopWav,self.buttonStop)
self.MainSizer.Add(self.buttonPlay,pos=(2,2),flag=wx.ALL,border=10)
self.MainSizer.Add(self.buttonPause,pos=(2,3),flag=wx.ALL,border=10)
self.MainSizer.Add(self.buttonStop,pos=(2,4),flag=wx.ALL,border=10)
self.slider=wx.Slider(self.annoPanel,-1,0,0,1)
self.slider.SetMinSize((150,-1))
self.MainSizer.Add(self.slider,pos=(3,2),span=(1,3),flag=wx.ALL|wx.EXPAND,border=10)
self.Bind(wx.EVT_SLIDER,self.OnSeek,self.slider)
self.timer=wx.Timer(self)
self.Bind(wx.EVT_TIMER,self.OnTimer)
self.timer.Start(100)
self.rec_st=wx.StaticText(self.annoPanel,label=u"语音识别结果")
self.rec_st.SetFont(font)
self.MainSizer.Add(self.rec_st,pos=(0,7),span=(1,4),flag=wx.EXPAND|wx.ALL,border=10)
self.MainSizer.AddGrowableCol(10,1)
self.rec_tc=wx.TextCtrl(self.annoPanel,value=self.recTextStr,style=wx.TE_READONLY|wx.TE_MULTILINE)
self.MainSizer.Add(self.rec_tc,pos=(2,7),span=(4,4),flag=wx.EXPAND|wx.ALL,border=10)
self.label_st=wx.StaticText(self.annoPanel,label=u"音频标记")
self.label_st.SetFont(font)
self.MainSizer.Add(self.label_st,pos=(0,12),flag=wx.ALL,border=10)
self.label_tc=wx.TextCtrl(self.annoPanel,style=wx.TE_MULTILINE)
self.MainSizer.Add(self.label_tc,pos=(2,12),span=(4,1),flag=wx.EXPAND|wx.ALL,border=10)
self.anno_st=wx.StaticText(self.annoPanel,label=u"文本标注")
self.anno_st.SetFont(font)
self.MainSizer.Add(self.anno_st,pos=(0,13),span=(1,4),flag=wx.EXPAND|wx.ALL,border=10)
self.MainSizer.AddGrowableCol(16,1)
self.anno_tc=wx.TextCtrl(self.annoPanel,style=wx.TE_MULTILINE)
self.MainSizer.Add(self.anno_tc,pos=(2,13),span=(4,4),flag=wx.EXPAND|wx.ALL,border=10)
self.line=wx.StaticLine(self.annoPanel)
self.MainSizer.Add(self.line,pos=(1,0),span=(1,18),flag=wx.EXPAND|wx.BOTTOM,border=10)
self.save_button=wx.Button(self.annoPanel,label=u"保存")
self.clear_button=wx.Button(self.annoPanel,label=u"清空")
self.MainSizer.Add(self.save_button,pos=(2,17),flag=wx.ALL,border=10)
self.MainSizer.Add(self.clear_button,pos=(3,17),flag=wx.ALL,border=10)
self.save_button.Bind(wx.EVT_BUTTON,self.ButtonSave)
self.clear_button.Bind(wx.EVT_BUTTON,self.ButtonClear)
self.InitMedia()
self.annoPanel.SetSizerAndFit(self.MainSizer)
#播放器进度条的定时器事件回调
def OnSeek(self, evt):
offset = self.slider.GetValue()
self.mc.Seek(offset)
def OnTimer(self, evt):
offset = self.mc.Tell()
self.slider.SetValue(offset)
#保存结果
def ButtonSave(self,event):
with codecs.open('anno_file','w','gbk') as f:
anno_text = self.anno_tc.GetValue()
print anno_text
f.write(anno_text)
self.list.SetStringItem(self.choosenItem,1,u"完成")
self.list.SetItemBackgroundColour(self.choosenItem,"green")
#删除中间操作的temp文件,并对最后保存的文件重命名
if len(self.tempwavlist) != 0:
lastfile=self.tempwavlist[-1]
self.tempwavlist.pop()
for i in self.tempwavlist:
print i
os.remove(i)
data = self.list.GetItem(self.choosenItem,0)
tuple_temp = self.wavList[int(data.GetText())-1]
print tuple_temp
path_str=os.path.split(tuple_temp[1])
wavname=path_str[1].split('.')
newname=wavname[0]+"_nocut."+wavname[1]
newpath=os.path.join(path_str[0],newname)
print newpath
os.rename(tuple_temp[1],newpath)
os.rename(lastfile,tuple_temp[1])
#清空标注文本框
def ButtonClear(self,event):
self.anno_tc.SetValue("")
self.label_tc.SetValue("")
#播放器播放
def playWav(self,event):
if not self.mc.Play():
wx.MessageBox("Unable to Play media : Unsupported format?","ERROR",wx.ICON_ERROR | wx.OK)
else:
print self.mc.Length()
self.slider.SetRange(0, self.mc.Length())
#播放器暂停
def pauseWav(self,event):
#if self.mc.Pause == 0:
# pygame.mixer.music.pause()
# self.buttonPause.SetLabel(u"▷")
# self.mc.Pause = 1
#else:
# pygame.mixer.music.unpause()
# self.buttonPause.SetLabel(u"‖")
# self.mc.Pause = 0
self.mc.Pause()
#播放器停止
def stopWav(self,event):
self.mc.Stop()
#--------------------------------------------------
#标签操作的UI创建
def OnLabelPaint(self,event):
brush=wx.Brush("white")
dc2=wx.PaintDC(self.labelPanel)
dc2.SetBackground(brush)
dc2.Clear()
dc2.DrawRectangle(79,30,1550,40)
for i in self.labelPos:
print i
dc4=wx.PaintDC(self.labelPanel)
dc4.DrawLine(i,30,i,70)
#--------------------------------------------------
#音频操作的UI创建
def CreateWavUI(self):
self.haveDraw = 0
self.haveCut = 0
self.wavpath_ST = wx.StaticText(self.cutPanel,-1,u"当前操作文件 : "+self.wavfilepath,pos=(40,10))
wx.StaticText(self.cutPanel,-1,u"时间 : ",pos=(40,60))
wx.StaticText(self.cutPanel,-1,u"坐标:",pos=(600,60))
self.labelButton=wx.Button(self.cutPanel,-1,"label",pos=(200,60))
self.cutButton=wx.Button(self.cutPanel,-1,"delete",pos=(300,60))
self.playButton=wx.Button(self.cutPanel,-1,"play",pos=(400,60))
self.posCtrl=wx.TextCtrl(self.cutPanel,-1,"",pos=(640,60))
self.timeCtrl=wx.TextCtrl(self.cutPanel,-1,"",pos=(80,60))
self.Bind(wx.EVT_BUTTON,self.OnLabelCreate,self.labelButton)
self.Bind(wx.EVT_BUTTON,self.cutWav,self.cutButton)
self.Bind(wx.EVT_BUTTON,self.PlayWav,self.playButton)
self.testbutton=wx.Button(self.cutPanel,-1,"revoke",pos=(500,60))
self.Bind(wx.EVT_BUTTON,self.testtest,self.testbutton)
self.recordBtn = wx.Button(self.cutPanel, -1, "Record", pos=(750, 61))
self.Bind(wx.EVT_BUTTON, self.OnPaint2, self.recordBtn)
def CreateUpWavUI(self):
self.uphaveDraw = 0
self.uphaveCut = 0
self.upwavpath_ST = wx.StaticText(self.upcutPanel,-1,u"当前操作文件 : "+self.wavfilepath,pos=(40,10))
wx.StaticText(self.upcutPanel,-1,u"时间 : ",pos=(40,60))
wx.StaticText(self.upcutPanel,-1,u"坐标:",pos=(600,60))
self.uplabelButton=wx.Button(self.upcutPanel,-1,"label",pos=(200,60))
self.upcutButton=wx.Button(self.upcutPanel,-1,"delete",pos=(300,60))
self.upplayButton=wx.Button(self.upcutPanel,-1,"play",pos=(400,60))
self.upposCtrl=wx.TextCtrl(self.upcutPanel,-1,"",pos=(640,60))
self.uptimeCtrl=wx.TextCtrl(self.upcutPanel,-1,"",pos=(80,60))
self.Bind(wx.EVT_BUTTON,self.OnLabelCreate,self.uplabelButton)
self.Bind(wx.EVT_BUTTON,self.cutWav,self.upcutButton)
self.Bind(wx.EVT_BUTTON,self.PlayWav,self.upplayButton)
self.uptestbutton=wx.Button(self.cutPanel,-1,"revoke",pos=(500,60))
self.Bind(wx.EVT_BUTTON,self.testtest,self.uptestbutton)
self.upecordBtn = wx.Button(self.cutPanel, -1, "Record", pos=(750, 61))
self.Bind(wx.EVT_BUTTON, self.OnPaint2, self.uprecordBtn)
#左键按下事件的回调函数,绘制一条线
def OnLeftDown(self,event):
self.wavPanel.Refresh()
self.firstPos=event.GetPosition()
#haveDraw为判断是否左键按下后是否有拖动动作
if self.haveDraw == 1:
self.haveDraw = 0
dc=wx.BufferedDC(wx.ClientDC(self.wavPanel))#,self.wavPanel
# dc=wx.BufferedPaintDC(self.wavPanel)
dc.DrawLine(self.firstPos.x, 30, self.firstPos.x, 270)
#左键松开事件的回调函数,绘制选中的矩形框
def OnDrawRec(self,event):
#haveDraw为判断是否左键按下后是否有拖动拖动动作
if self.haveDraw == 1:
dc=wx.BufferedDC(wx.ClientDC(self.wavPanel))
#dc=wx.BufferedPaintDC(self.wavPanel)
brush=wx.Brush(wx.Colour(255,23,140))
dc.SetBrush(brush)
dc.SetBrush(wx.Brush(wx.Colour(255,23,140),wx.BDIAGONAL_HATCH))
dc.DrawRectangle(self.firstPos.x,30,self.secondPos.x-self.firstPos.x,240)
else:
#dc=wx.PaintDC(self.wavPanel)
dc=wx.BufferedDC(wx.ClientDC(self.wavPanel))#,self.wavPanel
dc.DrawLine(self.firstPos.x,30,self.firstPos.x,270)
def OnMove(self, event):
pos=event.GetPosition()
self.posCtrl.SetValue("%s,%s" %(pos.x,pos.y))
#在左键按下和拖动下绘制选中的矩形框
if event.Dragging() and event.LeftIsDown():
self.secondPos=pos
self.wavPanel.Refresh()
dc = wx.BufferedDC(wx.ClientDC(self.wavPanel))
#dc=wx.BufferedPaintDC(self.wavPanel)
brush=wx.Brush(wx.Colour(100,100,100))
dc.SetBrush(brush)
dc.SetBrush(wx.Brush(wx.Colour(255,23,140),wx.BDIAGONAL_HATCH))
if self.firstPos.x<80 :
self.firstPos = 250
if self.firstPos.x>1630 :
self.firstPos = 1630
if self.firstPos.x<80 :
self.firstPos = 80
if self.firstPos.x>1630 :
self.firstPos = 1630
else:
dc.DrawRectangle(self.firstPos.x,30,self.secondPos.x-self.firstPos.x,240)
self.haveDraw = 1
#标签界面
def OnLabelMove(self,event):
pos=event.GetPosition()
#在按下左键拖动后绘制两条指示线
if event.Dragging() and event.LeftIsDown():
self.labelPanel.Refresh()
self.wavPanel.Refresh()
self.labelPosition=event.GetPosition()
a=self.wavTime*self.calPosition(pos.x,self.wavParams,self.wavLength,1)
self.timeCtrl.SetValue("%.3fs"%a)
dc=wx.ClientDC(self.labelPanel)
dc.DrawLine(self.labelPosition.x,30,self.labelPosition.x,70)
dc3=wx.BufferedDC(wx.ClientDC(self.wavPanel))
#dc3=wx.BufferedPaintDC(self.wavPanel)
dc3.DrawLine(self.labelPosition.x,30,self.labelPosition.x,270)
#松开左键后绘制两条指示线
def OnLabelDrawLine(self,event):
self.labeltime=self.timeCtrl.GetValue()
dc1 = wx.ClientDC(self.labelPanel)
dc1.DrawLine(self.labelPosition.x, 30, self.labelPosition.x, 70)
dc3 = wx.BufferedDC(wx.ClientDC(self.wavPanel))
#dc3=wx.BufferedPaintDC(self.wavPanel)
dc3.DrawLine(self.labelPosition.x, 30, self.labelPosition.x,270)
#创建输入标签的文本框
def OnLabelCreate(self,event):
self.labelPos.append(self.labelPosition.x)
self.labelPos.sort()
self.labelCtrl=wx.TextCtrl(self.labelPanel,-1,"",pos=(self.labelPosition.x-93,33),size=(90,33),style=wx.TE_PROCESS_ENTER)
self.labelCtrl.Bind(wx.EVT_TEXT_ENTER,self.TestEnter,self.labelCtrl)
self.labellist.append((str(self.labeltime),self.labelCtrl))
#标签文本框的enter建回调函数,若有值则保存,若无值按下回车则删除
def TestEnter(self,event):
self.labellist.sort()
notfind=0
enter_str=event.GetString()
label=event.GetEventObject()
for j in range(len(self.labellist)):
if label == self.labellist[j][1]:
labelindex=j
itemtime = self.labellist[j][0]
if enter_str == "":
for i in range(len(self.labellist)):
if label == self.labellist[i][1]:
del self.labellist[i]
del self.labelPos[i]
self.showlist.DeleteItem(i)
label.Destroy()
break
else:
if self.showlist.GetItemCount() != 0:
for i in range(self.showlist.GetItemCount()):
if self.showlist.GetItem(i,0).GetText() == itemtime:
self.showlist.SetStringItem(i, 1, label.GetValue())
notfind=1
break
if notfind == 0:
index = self.showlist.InsertStringItem(labelindex, itemtime)
self.showlist.SetStringItem(index, 1, label.GetValue())
else:
index = self.showlist.InsertStringItem(labelindex, str(self.labellist[0][0]))
self.showlist.SetStringItem(index, 1, self.labellist[0][1].GetValue())
self.bbbPanel.Refresh()
#创建标签显示的list界面,显示当前插入的标签的时间和内容
def CreateLabelListUI(self):
self.listbox=wx.BoxSizer(wx.HORIZONTAL)
self.showlist = wx.ListCtrl(self.bbbPanel, -1, style=wx.LC_REPORT|wx.LC_SORT_ASCENDING)#wx.LC_REPORT|
self.showlist.InsertColumn(0, u'时间', width=200)
self.showlist.InsertColumn(1, u'标签', width=410)
for i in self.labellist:
index = self.showlist.InsertStringItem(sys.maxint, str(i[0]))
label_str = i[1].GetValue()
print label_str
self.showlist.SetStringItem(index, 1, label_str)
self.listbox.Add(self.showlist,1,wx.EXPAND)
self.bbbPanel.SetSizerAndFit(self.listbox)
self.bbbPanel.Refresh()
#插销上一步的剪辑操作,实际为读取再前一次的缓存音频
def testtest(self,event):
print self.labellist
self.cutCount = self.cutCount - 2
if self.cutCount == 0:
self.cutCount = 1
data = self.list.GetItem(self.choosenItem,0)
tuple_temp = self.wavList[int(data.GetText())-1]
self.wavfilepath=tuple_temp[1]
if len(self.tempwavlist) != 0:
self.tempwavlist.pop()
print self.tempwavlist
# self.cutCount = self.cutCount + 1
self.drawBackJpg(self.wavfilepath)
self.InitMedia()
self.wavPanel.Refresh()
else:
print os.path.split(self.wavfilepath)
path_str=os.path.split(self.wavfilepath)
cutfile="temp"+str(self.cutCount)+".wav"
temppath=os.path.join(path_str[0],cutfile)
print temppath
self.cutCount = self.cutCount + 1
self.wavfilepath=temppath
self.tempwavlist.pop()
print self.tempwavlist
self.drawBackJpg(self.wavfilepath)
self.InitMedia()
self.wavPanel.Refresh()
#播放选中部分音频
def PlayWav(self,event):
f = wave.open(self.wavfilepath,"rb")
print "1.pos:",self.firstPos.x,"2.pos",self.secondPos.x
if self.firstPos.x < self.secondPos.x:
x1=self.firstPos.x
x2=self.secondPos.x
else:
x2=self.firstPos.x
x1=self.secondPos.x
print "x1:",x1,"x2",x2
params = f.getparams()
nchannels, sampwidth, framerate, nframes = params[:4]
str_data = f.readframes(nframes)
str_length = len(str_data)
print "str length = ",str_length
a=self.calPosition(x1,params,str_length,1)
b=self.calPosition(x2,params,str_length,1)
atime=round(a*self.wavTime,2)
btime=round(b*self.wavTime,2)
print atime,btime
alength=int(a*self.mc.Length())
print alength
self.mc.Seek(alength)
self.mc.Play()
time.sleep(btime-atime)
self.mc.Stop()
# str_test=str_data[a:b]
# print len(str_test)
# p = pyaudio.PyAudio()
# print f.getnchannels();
# print p.get_format_from_width(f.getsampwidth())
# print f.getframerate()
# stream = p.open(format = p.get_format_from_width(f.getsampwidth()),
# channels = f.getnchannels(),
# rate = f.getframerate(),
# output = True)
# print "==="
# stream.write(str_test)
# stream.stop_stream()
# stream.close()
# f.close()
# p.terminate()
#剪辑音频
def cutWav(self,event):
print self.wavfilepath
f = wave.open(self.wavfilepath,"rb")
print os.path.split(self.wavfilepath)
path_str=os.path.split(self.wavfilepath)
cutfile="temp"+str(self.cutCount)+".wav"
temppath=os.path.join(path_str[0],cutfile)
print ("debug: temppath:", temppath)
self.cutCount = self.cutCount + 1
self.tempwavlist.append(temppath)
print self.tempwavlist
self.wavfilepath = temppath
print "1.pos:",self.firstPos.x,"2.pos",self.secondPos.x
if self.firstPos.x < self.secondPos.x:
x1=self.firstPos.x
x2=self.secondPos.x
else:
x2=self.firstPos.x
x1=self.secondPos.x
print "x1:",x1,"x2",x2
params = f.getparams()
nchannels, sampwidth, framerate, nframes = params[:4]
str_data = f.readframes(nframes)
str_length = len(str_data)
print "str length",str_length
a=self.calPosition(x1,params,str_length,0)
b=self.calPosition(x2,params,str_length,0)
wf = wave.open(self.wavfilepath,"wb")
wf.setnchannels(nchannels)
wf.setsampwidth(sampwidth)
wf.setframerate(framerate)
str_test=str_data[0:a]+str_data[b:str_length]
print len(str_test)
wf.writeframes(str_test)
wf.close()
self.drawBackJpg(self.wavfilepath)
self.InitMedia()
self.wavPanel.Refresh()
#计算当前坐标的在音频中的相对位置
def calPosition(self,x,params,str_length,type):
nchannels, sampwidth, framerate, nframes = params[:4]
print nchannels, sampwidth, framerate, nframes
bytecount=int(nchannels*sampwidth)
print bytecount
real_length = int(((str_length//bytecount)/framerate)/((((str_length//bytecount)/framerate)//0.5+1)*0.5)*1550)
print real_length
aa=(x-79)/real_length
print aa
a = ((str_length//bytecount)//1000)*(int(((x-79)/real_length)*1000))*bytecount
print a
if type == 1:
# print "aa ",int(aa)
return aa
else:
# print "a",int(a)
return int(a)
#绘制波形图片
def drawBackJpg(self, path):
print ("debug path:", path)
f = wave.open(path, "r")
params = f.getparams()
nchannels, sampwidth, framerate, nframes = params[:4]
str_data = f.readframes(nframes)
f.close()
str_length = len(str_data)
self.wavLength = str_length
self.wavParams = params
time_limit=(((str_length//(nchannels*sampwidth)/framerate)//0.5+1)*0.5)
wave_data = numpy.fromstring(str_data,dtype = numpy.short)
wave_data.shape = -1, int(params[0])
wave_data = wave_data.T
time = numpy.arange(0, nframes)*(1.0/framerate)
len_time = len(time)
time = time[0:len_time]
self.wavTime = len(time)/16000
axes = pl.subplot(111)
axes.plot(time, wave_data[0], 'cornflowerblue')
axes.xaxis.set_ticks_position('top')
axes.xaxis.set_major_locator(xmajorLocator)
axes.xaxis.set_minor_locator(xminorLocator)
axes.xaxis.grid(True, which='minor')
axes.set_xlim([0.0, time_limit])
axes.set_yticks([])
fig = pl.gcf()
fig.set_size_inches(20, 3)
fig.savefig('mic4_output.png', dpi=100)
# gc.collect()
pl.close()
def record_and_drawBackJpg(self):
global input_filename
self.wavfilepath = input_filename
self.OnFreshUI()
self.drawBackJpg(self.wavfilepath)
# 重绘
brush = wx.Brush("white")
img = wx.Bitmap("mic4_output.png")
dc = wx.ClientDC(self.wavPanel)
dc.Clear()
dc.SetBackground(brush)
dc.DrawBitmap(img, -172, 0)
def OnPaint2(self, event):
# img = wx.Bitmap('res/times.png', wx.BITMAP_TYPE_ANY)
for i in range(0, int(11025 / 256 * 20)):
self.get_audio() # 实现一边录音,一边绘图
# 绘图
self.record_and_drawBackJpg()
def get_audio(self):
global in_path
global global_frames
CHUNK = 256
FORMAT = pyaudio.paInt16
CHANNELS = 1 # 声道数
RATE = 11025 # 采样率
RECORD_SECONDS = 2 # 每次录音 5s
WAVE_OUTPUT_FILENAME = in_path
p = pyaudio.PyAudio()
stream = p.open(format=FORMAT,
channels=CHANNELS,
rate=RATE,
input=True,
frames_per_buffer=CHUNK)
print("*"*10, u"start record: please speak in 5s")
frames = global_frames
for i in range(0, int(RATE / CHUNK * RECORD_SECONDS)):
data = stream.read(CHUNK)
frames.append(data)
print("*"*10, u"speak over\n")
global_frames = frames
stream.stop_stream()
stream.close()
p.terminate()
# ./input.wav
wf = wave.open(WAVE_OUTPUT_FILENAME, 'wb')
wf.setnchannels(CHANNELS)
wf.setsampwidth(p.get_sample_size(FORMAT))
wf.setframerate(RATE)
# 写wav
wf.writeframes(b''.join(global_frames))
wf.close()
# start
class LoginWin(wx.Frame):
def __init__(self,parent,title):
super(LoginWin, self).__init__(parent, title=title, size=(590,150))
self.InitUI()
self.Centre()
self.Show()
def InitUI(self):
self.LoginPanel = wx.Panel(self)
hbox = wx.BoxSizer(wx.HORIZONTAL)
fgs = wx.FlexGridSizer(2, 3, 10,10)
wavListST = wx.StaticText(self.LoginPanel, label = u"音频列表")
recTextST = wx.StaticText(self.LoginPanel, label = u"识别文本")
ChooseButton1 = wx.Button(self.LoginPanel, label=u"选择")
ChooseButton2 = wx.Button(self.LoginPanel, label=u"选择")
ChooseButton1.Bind(wx.EVT_BUTTON, self.ReadList)
ChooseButton2.Bind(wx.EVT_BUTTON, self.ReadText)
self.ListTC= wx.TextCtrl(self.LoginPanel, style=wx.TE_READONLY)
self.TextTC = wx.TextCtrl(self.LoginPanel, style=wx.TE_READONLY)
fgs.AddMany([(wavListST), (self.ListTC, 1, wx.EXPAND), (ChooseButton1), (recTextST),(self.TextTC, 1, wx.EXPAND), (ChooseButton2)])
fgs.AddGrowableCol(1, 1)
hbox.Add(fgs, proportion = 1, flag = wx.ALL|wx.EXPAND, border = 10)
OkButton = wx.Button(self.LoginPanel, label="OK")
CancelButton = wx.Button(self.LoginPanel, label="CANCEL")
OkButton.Bind(wx.EVT_BUTTON, self.OnClick)
sizer = wx.GridBagSizer(hgap=10, vgap=10)
sizer.Add(OkButton, pos=(0, 20), flag=wx.ALL)
sizer.Add(CancelButton, pos=(0, 21), flag=wx.ALL)
bsizer = wx.BoxSizer(wx.VERTICAL)
bsizer.Add(hbox,0,wx.EXPAND,10)
bsizer.Add(sizer,0,wx.EXPAND,10)
self.LoginPanel.SetSizerAndFit(bsizer)
sizer.SetSizeHints(self.LoginPanel)
def ReadList(self, event):
dlg = wx.FileDialog(self, message=u"选择音频列表文件select ",defaultDir=os.getcwd(),defaultFile="",style=wx.OPEN|wx.CHANGE_DIR)
self.wavList = ["mic4.wav"]
if dlg.ShowModal() == wx.ID_OK:
path = dlg.GetPath()
self.ListTC.SetValue(path)
print ("Debug: ", path)
# self.wavList = self.ReadFilePath(path)
dlg.Destroy()
def ReadText(self, event):
dlg = wx.FileDialog(self, message=u"选择识别结果文本",defaultDir=os.getcwd(), defaultFile="",style = wx.OPEN|wx.CHANGE_DIR)
self.recText = ["recog.txt"]
if dlg.ShowModal() == wx.ID_OK:
path = dlg.GetPath()
self.TextTC.SetValue(path)
# self.recText = se0lf.ReadFilePath(path)
print ("debug: rectext:", self.recText)
dlg.Destroy()
def ReadFilePath(self,path):
tempList = []
i = 1
with open(path, 'rb') as f:
for line in f.readlines():
tempList.append((i, line.strip('\n').decode('utf-8')))
i = i + 1
return tempList
def OnClick(self, event):
self.Close()
mainWin = MyFrame(self.wavList, self.recText)
mainWin.Show()
app=wx.App(0)
LoginWin(None,'login')
app.MainLoop()
本文章为转载内容，我们尊重原作者对文章享有的著作权。如有内容错误或侵权问题，欢迎原作者联系我们进行内容更正或删除文章。
