From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.
We appreciate your patience as we improve our online experience.
From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.
We appreciate your patience as we improve our online experience.
04-21-2014 11:10 AM
I want to perform 1d convolution without using any pre-defined Convolution VI's.I have been able to do it succesfully using zero-padding but now i want to implement it via basic method of shifting the array,then multiplying and summing.I have tried a lot but probably lost my way while performing shifting and multiplication and getting incorrect results.Pls suggest a way out to shift and multiply those arrays to perform convolution.
04-21-2014 11:24 AM
SInce LabVIEW ships wit convolution tools, this is a pure exercise and I recommend to try to complete it yourself. Looking at your code, you need to take a few LabVIEW tutorials first, because mostall of what you just did makes no sense at all.
Learn about autoindexing. Think why you thought you needed a "build array" creating a 2D from a 1D array without changing the total number of elements) before summing, etc. etc. Maybe learn about shift registers.
04-21-2014 11:26 AM - edited 04-21-2014 11:28 AM
Hi ps,
why don't you start with reading Wikipedia on convolution and then try to implement the basic formula given there?
This should be pretty easy… 🙂
04-21-2014 11:33 AM
GerdW wrote:why don't you start with reading Wikipedia on convolution and then try to implement the basic formula given there?
This should be pretty easy… 🙂
Since he talks about shifting and multiplying, I assume the poster is familiar with the math. what is lacking are the LabVIEW skills.
It was also mentioned that it works with "zoeropadding", which also makes no sense, because there is much more to it than just that.
04-21-2014 05:06 PM - edited 04-21-2014 05:06 PM
Here is the convolution code of one my very old example that gives the same result as the stock covolution.vi (within ~1e-15).
Try to understand it. 😉 There are many ways to do this, of course.