How to mask the data of no information?

Dear Matlab experts,

I need to mask out data with no information and want to keep data for only targets of interest.

I have attached one image in which data with no information is masked out and retain with target information.

I have attached .mat file, could you please guide on how to mask the information of no use and get only target region?

Thank you!

Best Regards,

Amjad

load('two_targets.mat'); % data
imagesc(intensity,[-45 0]);

I need to keep target area as given in image below and maskout information of no use for spatial mean operation.

I used to cut data or keep points of target but it affects spatial mean operation.

[Amjad Iqbal]

@Walter Roberson

could you please check this query?

Thank you!

[Walter Roberson]

What do you mean by "spatial mean" for this purpose?

What are the boundaries you want to mask with? Your .mat file does not contain anything that looks like a mask. Is the rule that two circles have to be constructed, both with center at half-height, with the two touching at half-width? If so then we would need to know the radius to use.

[Amjad Iqbal]

Dear @Walter Roberson

From .mat file we can have this image named *output* given below.

Now I need to keep information of these two circular targets, according to the colormap its value varies roughlt from -20 to 0

I want to keep this information and mask out rest of the information below -45 to -20 with NaN as given in image above

having white background and two targets.

After that I need to apply sliding window for spatial mean operation.

Answer 1

[Chunru]

websave("two_targets.mat", "https://www.mathworks.com/matlabcentral/answers/uploaded_files/1132000/Two_targets.mat");
load('two_targets.mat'); % data
imagesc(intensity,[-45 0]);

intensity(intensity<-20) = nan;
figure
imagesc(intensity,[-45 0]);

[Walter Roberson]

3 x 3 sliding window (sliding one pixel at a time)

websave("two_targets.mat", "https://www.mathworks.com/matlabcentral/answers/uploaded_files/1132000/Two_targets.mat");
load('two_targets.mat'); % data
imagesc(intensity,[-45 0]);

intensity(intensity<-20) = nan;

imagesc(intensity,[-45 0]);

[min(intensity(:)), max(intensity(:))]

ans = 1×2

-19.9535 0

smoothed = blockproc(intensity, [1 1], @(block) mean(block.data(:), 'omitnan'), 'Border', [1 1], 'TrimBorder', false);

imagesc(smoothed);

[min(smoothed(:)), max(smoothed(:))]

ans = 1×2

-19.9535 0

[Amjad Iqbal]

@Chunru and @Walter Roberson Thank you dear both!

It helps me a lot and work perfecly, and I learnt somthing new, I really appriciate your inputs