pcolor - how to smooth

Hi everyone~

I'm plotting some contour map (x, y, z)

If I use Origin, it does some nice smoothing.

In Maltab, using

pcolor(x, y, z); shading interp;

It shows all the noise.

See the picture please.<https://lh6.googleusercontent.com/-fqdFfNo1k5E/UiDzBmXPjtI/AAAAAAAAAz0/iSuJktzidb8/w1044-h352-no/plot.png>

The 3rd plot is a comparison: when data's not noisy, Matlab produces a flat background part (the blue part) too.

I wonder if there's anyway I can get do the same as in Origin using Matlab. Or do I have to smooth the data myself?

Any thoughts are appreciated. Thanks~

[Yuji Zhang]

picture is here: https://lh4.googleusercontent.com/-q3EXlgFJ84g/UiDMUkHHTCI/AAAAAAAAAzU/bOKUbmQMa9c/w1044-h442-no/plot.png

Thanks~

[Yuji Zhang]

Answer 1

[Kelly Kearney]

Pretty sure it's just a colormap difference, with origin setting everything less than -90 to black. Try this:

pcolor(x,y,z);
shading interp;
set(gca, 'clim', [-90 -32]);
colormap([0 0 0; jet]);
colorbar;

You'll probably have to fiddle a bit more with the colormap and color limits to get the exact same thing, but that should be close.

[Yuji Zhang]

Hi Kelly~

Thanks for your help. The method to change the low-background color is helpful~

The graph I provided was misleading. My question was actually: the low signal part (black part in 2nd graph) has noise. In the 1st graph the noise is all shown. I don't want it to show.

See the comparison: the 3rd graph is where there's no noise, the blue part is flat.

Please see the updated picture. Thank you~

Answer 2

[Image Analyst]

Can you somehow get an image out of that? Because images are a lot easier to smoother than a list of random coordinates.

[Yuji Zhang]

Hi~ Thank you for your help! Sure I got picture. The graph I uploaded is a picture.

I'm not familiar with imagine processing. I can't imagine smoothing a picture is easier than smoothing the data. Could you give me a link or something that I can learn how to smooth a picture from? Or could you just download the picture I provided and show me an example of smoothing(if it's easy and quick)? Thanks a lot~

[Image Analyst]

One way, though not necessarily the best way is to just use convolution to blur each color channel one at a time:

% Extract the individual red, green, and blue color channels.
redChannel = rgbImage(:, :, 1);
greenChannel = rgbImage(:, :, 2);
blueChannel = rgbImage(:, :, 3);

windowSize = 7;
kernel = ones(windowSize) / windowSize ^ 2;
% Blur the individual color channels.
smoothedPictureR = conv2(double(redChannel), kernel, 'same');
smoothedPictureG = conv2(double(greenChannel), kernel, 'same');
smoothedPictureB = conv2(double(blueChannel), kernel, 'same');

% Recombine separate color channels into a single, true color RGB image.
smoothrgbImage = uint8(cat(3, smoothedPictureR , smoothedPictureG, smoothedPictureB));
imshow(smoothrgbImage);

Answer 3

[Usenobong Akpan]

Hi,

Please I am a new user of matlab. i need help on the matlab code that would make me plot a single 3 axes graph. I am comparing power (W) and temp (degree celsius) over same time interval. I can if there were of the same unit i.e. W and W over same time interval; but here I am having difficulty.

pls can anyone help out?