Removing the background from imagesc image.

Hi everybody, I'd like to remove the dark blue component from imagesc figure without erasing the shades of litgh blu. In order to achieve something similir to this

Tihs is my code.

[datahist,xedges,yedges] = histcounts2(data(1,:), data(3,:), bins);

heatmap = datahist';

heatmap = interp2(heatmap,'cubic');

% Create the heatmap

imagesc(xedges, yedges, heatmap, 'CDataMapping', 'scaled');

ax = gca;

ax.YDir = 'normal';

colormap('jet');

Thank you in advance.

Answer 1

[Image Analyst]

You just need to either adjust the colormap, or overlay an RGB image onto your original image. See

How to overlay a foreground image over a background image? - MATLAB Answers - MATLAB Central

Image overlay using transparency | Steve on Image Processing

Please attach your image if you need more help and tell us the gray level at which you want the colormap to start. Something like

cmap = turbo(256);
% Start colors at 100.
cmap(1:101, 1) = 0:100;
cmap(1:101, 2) = 0:100;
cmap(1:101, 3) = 0:100;
colormap(cmap); % Apply the new colormap

[Riccardo Fiorotto]

Basically. The actually final image that I obtain using imagesc is the following one.

and I would like to remove the dark blu part to obtain an image similar to the next one.

I have tried to implement the following code:

figure(1)
bins = [binn{1,1}{1}, binn{1,1}{2}];

[datahist,xedges,yedges] = histcounts2(data(1,:), data(3,:), bins);


heatmap = datahist';
heatmap = interp2(heatmap,'cubic');

% Create the heatmap
imagesc(xedges, yedges, heatmap, 'CDataMapping', 'scaled');
ax = gca;
ax.YDir = 'normal';

jet = [1,1,1;jet];

colormap(jet);

xlabel("Time (us)")
ylabel("Sample Charge (nC)")

y_end = size(heatmap, 1);  % Coordinata y finale (altezza della heatmap)
for i = 1:length(p)
    line([p(i), p(i)], [-y_end, y_end], 'lineStyle','--', 'color','red');
end

But I obtained something like this:

As you can see there is an enormous difference between the second e the third image...

Have I explained my issue better?

[DGM]

It's not really clear what all the other stuff is in the example image, but if's clear that it's using a different color map and the slope of the data is opposite what you have. That is, the "background" region is represented by the colors at the top of the colormap (red-white) instead of the colors at the bottom (cyan-blue).

Either way, it's a different colormap.

Answer 2

[DGM]

What are your actual goals? What do you want in the areas you "removed"? Is the background actually an image? Is the foreground opaque? Is this for display only, or is the goal to produce a raster image?

This is one way.

% fake data
inpict = imread('cameraman.tif');
Z = peaks(256);

% create two axes
hax1 = axes;
hax2 = axes;

% display image in lower axes
imagesc(inpict,'parent',hax1);

% display data in upper axes
hi = imagesc(Z,'parent',hax2);

% define alpha based on the values in the data
% it's up to you to figure out what's appropriate
hi.AlphaData = Z >= 0.8;

% set colormaps
colormap(hax1,gray(256));
colormap(hax2,jet(256));

% wrangle axes setup
linkaxes([hax1 hax2]);
axis(hax1,'equal','tight'); 
set(hax2,'visible','off','xtick',[],'ytick',[]) % hide top axes decorations
set(hax2,'position',hax1.Position,'xlim',hax1.XLim,'ylim',hax1.YLim, ...
    'plotboxaspectratio',hax1.PlotBoxAspectRatio); % try to align both axes

This relies on overlaying multiple axes in a figure. This assumes the goal is:

• background is an image
• foreground is opaque
• masking is binary
• undecorated raster output is not needed

If you want an actual raster image without dealing with figure wrangling, aliasing damage, etc, then this is how I'd do it in MIMT. This could be done without MIMT, but conveniences exist for a reason.

% fake data
inpict = imread('cameraman.tif');
Z = peaks(256);

% define alpha based on the values in the data
% this is a smooth PWL eased thresholding operation
w = 0.3; % ease width (data units)
c = 0.8; % threshold center (data units)
FGa = imclamp((Z - c)/w + 0.5);

% convert data to pseudocolor RGB image
FG = gray2pcolor(Z,jet(256),'cdscale');

% composite the two images
outpict = replacepixels(FG,inpict,FGa);

imshow(outpict)

Answer 3

[DGM]

Different colormap examples:

Z = im2double(imread('bbbb.png'));

% background is at the bottom of the map
imagesc(Z)
colormap(cwjet(256))
colorbar

% background is at the top of the map
imagesc(1-Z)
colormap(hwjet(256))
colorbar