The SAR data is complex in nature. the Complex numbers are composed of real and imaginary parts. These are stored in two files (i.e layers) which are indicated as i and q where i is the real and q is the imaginary. You don't need to deal with these two layers directly as SNAP is generated the intensity image from these two layers.

I want to add that you can derive the amplitude and the phase from the i,q layers.

The amplitude = sqrt(i^2 + q^2)

The phase = arctan (i / q)

Thanks for your answer Ahmed. Could you please let me know whether the data that I imported in SNAP toolbox has I and Q data with in it or not? Because the Amplitude band is created on-the-fly with in the toolbox and is not saved in the data imported.

Hi Pavan,

When you import the SLC image to SNAP, you will have all the layers you need to work with. As specified in the image that you attached, you have the real and the imaginary components of the image, as well as the phase and the amplitude. I am not sure what you want to do next. Please give me more details about that so I can help you with the next step.

Best Regards

Hi Ahmed,

Now that you have confirmed me with my assumption that the data imported in SNAP has I and Q information with it, the next step would be to display the images using MATLAB in RGB .

Thanks

Best regards

Pavan.

Hi Pavan,

If you want to read the files in Matlab, you should be familiar with the file structure. As I understand, it is in big tiff format. The data for each swath is stored in a single file, where they stored the pixel information as two seccessive numbers (real then imaginary). So, you have to separate the real and imaginary parts of that swath into two files then you can generate the amplitude and the phase components of that swath. It is a complicated process and I wish you the good luck with it.

Ahmed Athab

Hi Ahmed,

I found the format of the file structure of the data file. And was able to display the images using MATLAB. But now I am having a problem.

As you have mentioned that the pixel information is stored as two successive numbers representing Real and Imaginary values. Because of which the size of the data is halved each (for real and imaginary). For example, the size of the data I have with me is 7863 X 2120 and as I am separating the data into real and imaginary values, the size is reduced to half i.e. 7863 X 1060 each. Do you have any idea why is this the case?

But SNAP is showing me that size of the data is same for both I and Q for each polarisation i.e. 7863 X 2120.

Thanks in advance.

Best regards

Pavan.

Hi Pavan,

If SNAP shows the real and imaginary layers as (7863 (lines) x 2120 (columns)), then the original file size (mixed) should be (7863, lines x (2x2120,) columns) . Where each pixel in the row has two numbers.

When you read the file in Matlab into real and imaginary matrices, provided that you know the file dimensions, what you get is two matrecis, both are equal in size, having the same number of lines as the original file but half the number of columns. If you need more explination, please let me know.

Regards

Ahmed Athab

Hi Ahmed,

Thanks for a detailed explanation. I found it later after posting it over here. Yes the data format is 7863 X 2120 X 2.

Do you have any idea how to read a CEOS data file in MATLAB? I am having much tough time with this format of data.

Thank you.

Kind regards

Pavan.

I didn't work with such data. If you like, send me a sample file and I will try to load it to Matlab if you want.

Here is the link for the data which I am referring to.

ALOS2: http://seom.esa.int/polarimetrycourse2017/files/materials/SAN_FRANCISCO_ALOS2.zip

Format of the data: http://www.eorc.jaxa.jp/ALOS-2/en/doc/fdata/PALSAR-2_xx_Format_CEOS_E_f.pdf

Thanks in advance.

Ahmed Dhahir Athab , regarding your response about phase angle, the ecuation is

phase = arctan (q / i)