1
$\begingroup$

I have two processes, one is normal and the other with some cracks of material. And I have the acoustic signals got from the materials by a sensor. I want to compare these signals to see if this signal can be used to detect cracks. I think two different processes (with and without cracks) will create two different signals. Based on that, I want to compare these signals, but I'm new to DSP.

Could you please give me some suggestions what I should do, step by step to process these signals?

(I will of course read books to get knowledge about DSP, but it's really huge. I therefore hope that someone can tell me what I should do, step by step to process those signals. By doing so, I can save some time)

$\endgroup$

migrated from stackoverflow.com Sep 2 '12 at 23:32

This question came from our site for professional and enthusiast programmers.

  • 2
    $\begingroup$ What are you hoping to learn from the comparison? $\endgroup$ – datageist Sep 2 '12 at 23:34
  • $\begingroup$ What do you mean "I have non-stationary audio signals from 2 different processes" precisely? I did not understand what did you mean by 'process'. $\endgroup$ – Deniz Sep 3 '12 at 0:14
  • $\begingroup$ I have two processes, one is normal and the other with some cracks of material. And I have the acoustic signals got from the materials by a sensor. I want to compare these signals to see if this signal can be used to detect cracks. I have read somethings about DSP but it's really huge and it's not my major. I therefore hope that someone can give me some suggestions what I should do, what I should read. It seems to be a stupid question, but I hope I can make it faster with your help. Thank you. $\endgroup$ – John Sep 3 '12 at 9:09
  • 1
    $\begingroup$ So, you measured two audio signal from materials. Then, you have a 'normal' signal which can be thought as ground truth. And you have an 'abnormal' signal which characterizes cracks or other problems in material. You want to detect crack material by audio signal by using 'normal' signal. Am I right so far? $\endgroup$ – Deniz Sep 3 '12 at 10:22
  • 1
    $\begingroup$ @John I updated your question with some of the detail you provided in the comments. It would be helpful to provide some information about how the signals were generated (e.g. Is this from machinery that's already vibrating, or did you have to hit it with a hammer to generate the signal?). $\endgroup$ – datageist Sep 3 '12 at 17:29
1
$\begingroup$

Not really a programming question!

In any case if you need to read books to understand DSP and are trying to compare two non stationary processes I suggest you start reading right away because most steps I tell you won't make sense and you need to answer a host of questions even before anyone can answer this question.

Signal processing deals with lot of statistical phenomena and you have to work with it emperically many a times. Not to say there are no fixed rules but the step-by-step to process this signals is hard enough when data is stationary. Given that you have non stationary data, get reading and once you have a hang of it ask some pointed questions on dsp.stackexchange.com

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.