LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Adding a chaotic signal to an image

Hello there,

 I  want to add a signal (xm) to an image and after doing this subtract another signal (xs) from this summation. After this process I want also observe the image both after adding and subtracting and I could not be able to do all these things. I have tried to use image processing operators which is in Vision& Motion palette and later on I have found out that they are only for changing colors of desired image , besides connecting a signal to constant port is meaningless... I put here a screenshot which shows the work I want to do. I hope someone can help me about that issue... (Please note that xm and xs are chaotic signals.)

Thanks..

Regards..

Download All
0 Kudos
Message 1 of 10
(2,925 Views)

Hi gunaybulut,

 

I have a few additional questions:

1. what do your signals look like? Because, if you wire them like you do in your example, it is a single value, which is added to each point, making each point xm brighter or xs darker.

2. Currently, you are kinda doing "everything" at once. Especially, when loading an image and manipulating it, this becomes a problem. I think due to the reference based nature of IMAQ, you might get lucky, but I'd still bring it into some kind of data flow.

3. Can you describe your task simpler, maybe with an example? The input image, signal, and expected result would help things along quite a bit.



Remember Cunningham's Law
0 Kudos
Message 2 of 10
(2,892 Views)

@PeterFoerster asked the key question -- what do you mean by "signal"?  Is it the image, itself?  Are you trying to "blur" or "obscure" it?  Or does the Image represent (or "show") the "signal", like an image of an oscilloscope trace, to which you want to add "noise" (to the trace, not to the image of the trace) and subtract something?

 

The first problem is fairly easy, particularly if you are working with grey-scale images (if the images are color, then the question becomes "do you want to change the color, the intensity, or something else?"), but you, again, have to define the question a little better.  A picture of your "image" (and, perhaps, the "image" you wish to subtract) would also be helpful.

 

Bob Schor

0 Kudos
Message 3 of 10
(2,885 Views)

1. Plottings of xm and xs signals are in the attachment.

2. Well I have to synchronize two chaotic systems (already did) and make image encryption at the same time. I think that means I have to do things simultaneously. That's why I want to observe images after addition and subtraction to see that the image is modulated and demodulated by the chaotic signals.

3.I think I have already describe my task in article 2. But I still put a diagram on image encryption here to explain what I want to do (still not sure to be able to realize that task in LabVIEW).

 

Thanks...

Regards..

0 Kudos
Message 4 of 10
(2,884 Views)

I am currently trying to encrypt the image by blurring and also my image is a color image. The image which I used is also in the attachment.

 

 

Thanks..

Regards..

0 Kudos
Message 5 of 10
(2,882 Views)

I'll answer in more detail later. But here's my two cents:

 

Parallelization is a good concept for many things. It does not, however, apply to everything. (IMHO, it applies to about 5% of the cases where you expect it). E.g. you cannot parallelize getting into your car, getting out of your parking lot and driving to the supermarket. Those are inherently serialized tasks.



Remember Cunningham's Law
0 Kudos
Message 6 of 10
(2,861 Views)

Thank you for showing the "signals".  Of course, now I am more confused -- you appear to have a perfectly good Time Series (or perhaps even a LabVIEW Waveform), which you have represented as a 2D Image (why is unclear, but maybe you don't have access to the "raw data", but only have this 2D representation).

 

If XM and XS are represented by a Waveform, then the operations you want to do (encrypted transmission, decryption, addition/subtraction) are simple.  Furthermore, you can do other possibly-meaningful operations on the data, such as getting power spectra, counting peaks, getting statistical measures, etc. that are much more "natural" than operating on a 2D image of a time-series.

 

Do you have access to the original time series data?  If you do not, does it make (more) sense to try to extract the Time Series from the Image, and then do everything with the Time Series?

 

Bob Schor

0 Kudos
Message 7 of 10
(2,832 Views)

Well, I am also confused right now , but the thing is I do not know which block I need to use to convert the image into suitable data type. I have added the screenshot of my VI. That's the all data I have as you can see , I do not have any other data but only  have chaotic signal state variables data that it is plotted for 30 seconds according to time. I need to add these data which is sampled for 30 seconds to the image and subtract another same kind of data after this summation. The problem is I cannot add them directly cuz they are different type of datas and need to learn how I can convert the image to this type of data for to be able to adding them. ( In addition, I want to observe the image while doing these subtraction/ addition operations to see if it is  encrypted.)

0 Kudos
Message 8 of 10
(2,814 Views)

Now I'm getting seriously worried.  It seems clear that you do not have access to the data that form the basis for the Images you have posted, but only to pictures of the data.  Furthermore, you seem to be enlisting our help to encrypt (= conceal?) these images of data.  I don't feel comfortable providing much technical help for what seems to be a questionable enterprise ...

 

Bob Schor

0 Kudos
Message 9 of 10
(2,778 Views)

@Bob_Schor wrote:

Now I'm getting seriously worried.  It seems clear that you do not have access to the data that form the basis for the Images you have posted, but only to pictures of the data.  Furthermore, you seem to be enlisting our help to encrypt (= conceal?) these images of data.  I don't feel comfortable providing much technical help for what seems to be a questionable enterprise ...

 

Bob Schor


 

https://en.wikipedia.org/wiki/Steganography

0 Kudos
Message 10 of 10
(2,753 Views)