如何从 MATLAB 中的自相关数据中提取峰值?
我有关于音轨的信息(20,000 帧数据),我使用它自动关联:
[r,lags] = xcorr(XX,XX,'biased');
它看起来像这样:
希望到目前为止一切顺利。理想情况下,我希望能够获取与第二个峰值的最高部分相对应的帧号。我已经阅读并尝试了很多不同的方法,但我似乎无法让它为我检索信息。
有人能够阐明我必须做什么吗?
非常感谢!
编辑1: 我尝试过使用 findpeaks,但它似乎对我不起作用。我不确定这是否是因为我使用了错误的数据。
edit2:我目前正在测试一种仅在该音轨上使用的方法,但很快我想扩展它,以便我可以在整个文件目录上执行此方法,所以我需要一个可以检测峰值而不是自己查找信息的脚本。
edit3:我的.M文件:
[y, fs, nb] = wavread('Three.wav'); %# Load the signal into variable y
frameWidth = 441; %# 10ms
numSamples = length(y); %# Number of samples in y
numFrames = floor(numSamples/frameWidth); %# Number of full frames in y
energy = zeros(1,numFrames); %# Initialize energy
startSample = zeros(1,numFrames); %# Initialize start indices
endSample = zeros(1,numFrames); %# Initialize end indices
for frame = 1:numFrames %# Loop over frames
startSample(frame) = (frame-1)*frameWidth+1; %# Starting index of frame
endSample(frame) = frame*frameWidth; %# Ending index of frame
frameIndex = startSample(frame):endSample(frame); %# Indices of frame samples
energy(frame) = sum(y(frameIndex).^2); %# Calculate frame energy
end %# End loop
XX = filtfilt(ones(1,10)/10, 1, energy); %# Smooths signal
[r,lags] = xcorr(XX,XX,'biased'); %# Auto-correlates the data
plot(lags,r), xlabel('lag'), ylabel('xcorr') %# Plots data
I have information (20,000 frames of data) about an audio track that I have auto-correlated using:
[r,lags] = xcorr(XX,XX,'biased');
And it looks like this:
Which hopefully is so far so good. Ideally I would like to be able to take the frame number that corresponds to the highest part of the second peak. I've read around and tried a load of different methods but I just can't seem to get it to retrieve the information for me.
Would anybody be able to shed some light on what I have to do?
Many thanks!
edit1:
I have tried using findpeaks
, but it doesn't seem to work for me. I'm not sure if that's because I'm using the wrong data or not.
edit2: I'm currently testing a method to use on just this audio track, but soon I want to expand it so that I can perform this method on a whole directory of files, so I kind of need a script that can detect peaks rather than finding the information myself.
edit3: My .M file:
[y, fs, nb] = wavread('Three.wav'); %# Load the signal into variable y
frameWidth = 441; %# 10ms
numSamples = length(y); %# Number of samples in y
numFrames = floor(numSamples/frameWidth); %# Number of full frames in y
energy = zeros(1,numFrames); %# Initialize energy
startSample = zeros(1,numFrames); %# Initialize start indices
endSample = zeros(1,numFrames); %# Initialize end indices
for frame = 1:numFrames %# Loop over frames
startSample(frame) = (frame-1)*frameWidth+1; %# Starting index of frame
endSample(frame) = frame*frameWidth; %# Ending index of frame
frameIndex = startSample(frame):endSample(frame); %# Indices of frame samples
energy(frame) = sum(y(frameIndex).^2); %# Calculate frame energy
end %# End loop
XX = filtfilt(ones(1,10)/10, 1, energy); %# Smooths signal
[r,lags] = xcorr(XX,XX,'biased'); %# Auto-correlates the data
plot(lags,r), xlabel('lag'), ylabel('xcorr') %# Plots data
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(5)
编辑:
以及与标记峰值相对应的滞后值:
请注意,我们平滑了在计算自相关函数的导数之前计算信号,以便找到极值点
EDIT:
and the lag values corresponding to the marked peaks:
Note that we smoothed the signal prior to computing the derivative of the autocorrelation function in order to find the extrema points
如果您有信号处理工具箱,我认为 findpeaks 函数应该可以为您完成这项工作。
类似于:
http://www.mathworks.com/access /helpdesk/help/toolbox/signal/findpeaks.html
我习惯于在图像强度数据上使用它,图像强度数据的局部方差不如您的数据(那些看起来“厚”的部分)绘图只是许多快速上升和下降的数据点,对吧?)。您可能需要先对数据进行一些平滑处理才能使其正常工作。
If you have the signal processing toolbox, I think the findpeaks function should do the job for you.
Something like:
http://www.mathworks.com/access/helpdesk/help/toolbox/signal/findpeaks.html
I'm used to using it on image intensity data, which doesn't have quite as much local variance as your data (Those "thick" looking sections of your plot are just many data points going up and down quickly, right?). You might need to smooth the data out a bit first to get it to work.
第一步,您应该使用 xcorr 的第二个输出参数来使绘图上的单位正确:
现在您可以使用您最喜欢的方法来获取第二个峰值的滞后;如果您只需要执行一次,则图中的数据浏览器按钮即可。如果您需要重复执行此操作,我看不出有任何方法可以避免陷入
findpeaks
或类似情况。As a first step, you should use the second output argument of
xcorr
to get your units on the plot correct:Now you can use your favorite method to get the lag for the second peak; the data explorer button in the figure will do if you just need to do it once. If you need to do it repeatedly, I don't see any way to avoid getting into
findpeaks
or similar.您可以使用 findpeaks 两次。首先获得峰值的初始估计,并使用该结果来微调第二次调用 findpeaks 的输入参数。从问题来看,您似乎想计算音高值。
代码如下:
上面是此处给出的代码的修改版本。
You can use findpeaks twice. First to get an initial estimate of the peaks and use that results to fine tune the input parameters of the second call to findpeaks. From the question it seems that you want to calculate the pitch value.
Here is the code :
The above is a modified version of the code given here.