Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

Thread Subject:
edge detection

Subject: edge detection

From: Armindo

Date: 11 Jul, 2011 20:23:07

Message: 1 of 8

Hi

Please can anyone help me on this:

Let say that I have several values around - 9.81 and than the curve rise for positive values. I need to detect de value and indice at the edge when the signal initiate the rising.

There is some function that I can use to detect this?

I really need to be able to do this because I have too many files with data and this is not possible to do manually.

Any help would be apreciated.

Subject: edge detection

From: ImageAnalyst

Date: 11 Jul, 2011 23:14:15

Message: 2 of 8

There is an edge function in the Image Processing Toolbox, or you
could use conv() or conv2() with an appropriate kernel, or you could
use thresholding and find() like this
highDataIndexes = find(data > 9.81);

Subject: edge detection

From: Rune Allnor

Date: 12 Jul, 2011 07:08:26

Message: 3 of 8

On Jul 11, 10:23 pm, "Armindo " <armindom...@gmail.com> wrote:
> Hi
>
> Please can anyone help me on this:
>
> Let say that I have several values around - 9.81 and than the curve rise for positive values. I need to detect de value and indice at the edge when the signal initiate the rising.
>
> There is some function that I can use to detect this?
>
> I really need to be able to do this because I have too many files with data and this is not possible to do manually.
>
> Any help would be apreciated.

Read the books by van Trees on Detection and Estimation.

Rune

Subject: edge detection

From: Yumnam Kirani

Date: 12 Jul, 2011 12:01:09

Message: 4 of 8

ImageAnalyst <imageanalyst@mailinator.com> wrote in message <c13e8743-b5fa-483a-964d-4d6d33a0a793@q1g2000vbj.googlegroups.com>...
> There is an edge function in the Image Processing Toolbox, or you
> could use conv() or conv2() with an appropriate kernel, or you could
> use thresholding and find() like this
> highDataIndexes = find(data > 9.81);

You can use imfilter function. Using conv2 function increases the resulting image size, depending on the size of kernal or filter you used.

Yumnam Kirani Singh
Tronglaobi Awang Leikai

Subject: edge detection

From: Armindo

Date: 12 Jul, 2011 15:54:09

Message: 5 of 8

"Yumnam Kirani" <kirani.singh@gmail.com> wrote in message <ivhd25$igd$1@newscl01ah.mathworks.com>...
> ImageAnalyst <imageanalyst@mailinator.com> wrote in message <c13e8743-b5fa-483a-964d-4d6d33a0a793@q1g2000vbj.googlegroups.com>...
> > There is an edge function in the Image Processing Toolbox, or you
> > could use conv() or conv2() with an appropriate kernel, or you could
> > use thresholding and find() like this
> > highDataIndexes = find(data > 9.81);
>
> You can use imfilter function. Using conv2 function increases the resulting image size, depending on the size of kernal or filter you used.
>
> Yumnam Kirani Singh
> Tronglaobi Awang Leikai


Thanks for the help but you can explain a litle bit more how to tho this?

I try to use the highDataIndexes = find(data > 9.81); but this dont worke very well and detect points out of place.

I dont have an image but Acelaration values over time. And I need to detect when the aceleration (around - 9.81 it oscilates a litle bit) rise to positive values. I even saw this algorithm
http://www.cyut.edu.tw/~yltang/course/image%20processing/harris.m
but i have not been able to put this to work.

Please can anyone help me on this I really need to solve this problem.

I have a upload (link below) the a mat file with data over time and a picture showing the corners that I need to detect.

http://ifile.it/y4z5c17/Values.mat
http://ifile.it/o2svril/Values.JPG

kind regards

Subject: edge detection

From: Florin Neacsu

Date: 12 Jul, 2011 17:22:09

Message: 6 of 8

"Yumnam Kirani" <kirani.singh@gmail.com> wrote in message <ivhd25$igd$1@newscl01ah.mathworks.com>...
> ImageAnalyst <imageanalyst@mailinator.com> wrote in message <c13e8743-b5fa-483a-964d-4d6d33a0a793@q1g2000vbj.googlegroups.com>...
> > There is an edge function in the Image Processing Toolbox, or you
> > could use conv() or conv2() with an appropriate kernel, or you could
> > use thresholding and find() like this
> > highDataIndexes = find(data > 9.81);
>
> You can use imfilter function. Using conv2 function increases the resulting image size, depending on the size of kernal or filter you used.
>
> Yumnam Kirani Singh
> Tronglaobi Awang Leikai

Hello,

Would you please explain :
> You can use imfilter function. Using conv2 function increases the resulting image size, depending on the size of kernal or filter you used.

I disagree with this.

Regards,
Florin

Subject: edge detection

From: ImageAnalyst

Date: 12 Jul, 2011 18:57:33

Message: 7 of 8

On Jul 12, 1:22 pm, "Florin Neacsu" <fneac...@gmail.com> wrote:
> Hello,
>
> Would you please explain :
>
> > You can use imfilter function. Using conv2 function increases the resulting image size, depending on the size of kernal or filter you used.
>
> I disagree with this.
>
> Regards,
> Florin
------------------------------------------------------------------------
You can use imfilter to get the same result as conv2. Though, I don't
know if imfilter is as optimized for speed as conv2 (it might be).
Both imfilter and conv2 have the same output options - they can either
enlarge the image, or crop it to return an image the same size as the
input image, so there is no difference between the two functions
there.
ImageAnalyst

Subject: edge detection

From: Florin Neacsu

Date: 12 Jul, 2011 19:10:15

Message: 8 of 8

ImageAnalyst <imageanalyst@mailinator.com> wrote in message <dc402ef9-96a6-4083-8cc8-1eb10cface52@h12g2000vbx.googlegroups.com>...
> On Jul 12, 1:22 pm, "Florin Neacsu" <fneac...@gmail.com> wrote:
> > Hello,
> >
> > Would you please explain :
> >
> > > You can use imfilter function. Using conv2 function increases the resulting image size, depending on the size of kernal or filter you used.
> >
> > I disagree with this.
> >
> > Regards,
> > Florin
> ------------------------------------------------------------------------
> You can use imfilter to get the same result as conv2. Though, I don't
> know if imfilter is as optimized for speed as conv2 (it might be).
> Both imfilter and conv2 have the same output options - they can either
> enlarge the image, or crop it to return an image the same size as the
> input image, so there is no difference between the two functions
> there.
> ImageAnalyst

Hi ImageAnalyst,

That is exactly what I thought. Also conv2 is a standard Matlab function, whereas imfilter is part of the Image processing toolbox. Anyway, I was confused why Yumnam disconsidered conv2 so "strongly".

Regards,
Florin

Tags for this Thread

No tags are associated with this thread.

What are tags?

A tag is like a keyword or category label associated with each thread. Tags make it easier for you to find threads of interest.

Anyone can tag a thread. Tags are public and visible to everyone.

Contact us