Filtering the images

Share icon

Tue, 07/17/2012 - 07:13 am


Dear all,

I want to filter images using the Igor pro software. What codes do I use to filter an image such as fig.1 to obtain an image such as fig.2 ? I have just started using this software and I don't know much about it!

 Figures.pdf
Hello Arlette.

It is difficult to tell that the two images you show represent an image and its filtered version. The differences in the vertical axis and the color scale suggest otherwise.

Normally when you filter an image you have a specific goal in mind, e.g., noise reduction, enhancement, etc. It would be useful if you told us more about your requirements. You can send an IGOR experiment containing sample images to support@wavemetrics.com and we'll try to help you determine how to proceed.

A.G.
WaveMetrics, Inc.

July 17, 2012 at 12:45 pm - Permalink

arlette wrote:
I have just started using this software and I don't know much about it!


If you have not done it already, please go through the first half of the Igor Guided Tour by choosing Help->Getting Started. This is essential for learning Igor.

To learn about image processing, execute:
DisplayHelpTopic "Image Processing"

July 17, 2012 at 01:15 pm - Permalink

Dear A.G. WaveMetrics, Inc,

Thanks for the prompt response to my request for assistance. My problem is: I have a set of data ("first set of data" attached) that I need to enhance (remove the noise and/or grid in spectra taken in snapshot mode) in order for me to construct the Fermi surface. I want to enhance the data by removing all the noise and/or grids.

How do I do that, please ?



Many thanks,

Arlette Ngankeu
first set of data.zip

July 18, 2012 at 04:27 am - Permalink

I just had a quick look at "firstlayer" and "secondlayer". Some individual pixels seem to be out of range compared to the other ones.
The command:

NewImage firstlayer  


produces a black image because the highest value is scaled to white and zero to black (grey values in-between; most pixels are closer to zero compared to the max value).

A command like
firstlayer = firstlayer > 500 ? NAN : firstlayer


would remove the possibly flawed pixels and you would get something with looks like an image. The value of 500 is quite arbitrary - I guess there are constraints from the detector what the max value can be. NaN (the replacement value) is also arbitrary.

But I'm not sure if this is already what you mean by removing noise....


C

July 18, 2012 at 06:32 am - Permalink


Dear ChrLie,

Thank you very much, the command "firstlayer = firstlayer > 500 ? NAN : firstlayer" was very helpful and I finally construct the Fermi surface.

Arlette.

July 20, 2012 at 12:58 am - Permalink