2D colormaps. MxN matrix of RGB values for 4 colors gradient

c{1} = [1 0 0]; % specify 4 colors
c{2} = [0 1 0];
c{3} = [0 0 1];
c{4} = [1 0.5 0.2];
C = reshape(vertcat(c{:}), 2, 2, []);
n = 700; % number of pixels
img = zeros(n, n, 3);
for i=1:3
    img(:,:,i) = interp2([0 1], [0 1], C(:,:,i), linspace(0,1,n), linspace(0,1,n).');
end
imshow(img);

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Felipe Bayona 2020-9-17

Ameer

Thx for your help.

The clue is this interp2 function I didn't know. Works perfect

请先登录，再进行评论。

Answer 2

Bjorn Gustavsson 2020-9-17

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/595354-2d-colormaps-mxn-matrix-of-rgb-values-for-4-colors-gradient#answer_496162

If you do something like this:

[hsvImg] = rgb2hsv(Imrgb);
x_lims = [1 184;
          199 383;
          397 581;
          593 779;
          791 977;
          991 size(Imrgb,2)];
for i1 = 6:-1:1,
  subplot(4,6,i1)
  imagesc(Imrgb(:,x_lims(i1,1):x_lims(i1,2),:))
  subplot(4,6,i1+6*1)
  imagesc(hsvImg(:,x_lims(i1,1):x_lims(i1,2),1))
  subplot(4,6,i1+6*2)
  imagesc(hsvImg(:,x_lims(i1,1):x_lims(i1,2),2))
  subplot(4,6,i1+6*3)
  imagesc(hsvImg(:,x_lims(i1,1):x_lims(i1,2),3))
end
subplot(4,6,1)
ylabel('RGB-Images')
subplot(4,6,1+6*1)
ylabel('Hue')
subplot(4,6,1+6*2)
ylabel('Saturation')
subplot(4,6,1+6*3)
ylabel('Intensity')
figure
for i1 = 6:-1:1,
  subplot(4,6,i1)
  imagesc(Imrgb(:,x_lims(i1,1):x_lims(i1,2),:))
  subplot(4,6,i1+6*1)
  imagesc(Imrgb(:,x_lims(i1,1):x_lims(i1,2),1))
  subplot(4,6,i1+6*2)
  imagesc(Imrgb(:,x_lims(i1,1):x_lims(i1,2),2))
  subplot(4,6,i1+6*3)
  imagesc(Imrgb(:,x_lims(i1,1):x_lims(i1,2),3))
end
subplot(4,6,1)
ylabel('RGB-Images')
subplot(4,6,1+6*1)
ylabel('Red')
subplot(4,6,1+6*2)
ylabel('Green')
subplot(4,6,1+6*3)
ylabel('Blue')

You get to look at the Hue, Saturation and Intensity variation of the 6 different sub-images, and their respective red, green and blue image-planes. From there you should see that some of them are "reasonably" simple, and you should be able to reproduce them.

For mapping 2 data-sets, lets say I1 and I2, it should simplify to at most three 2-D interpolations, perhaps something like this:

I1_linear = linspace(min(I1(:)),max(I1(:)),suitable_nr4size1);
I2_linear = linspace(min(I2(:)),max(I2(:)),suitable_nr4size2);
hsvI2I2(:,:,3) = interp2(I1_linear,I2_linear,hsvImg(:,x_lims(i1,1):x_lims(i1,2),3),I1,I2);
hsvI2I2(:,:,2) = interp2(I1_linear,I2_linear,hsvImg(:,x_lims(i1,1):x_lims(i1,2),2),I1,I2);
hsvI2I2(:,:,2) = interp2(I1_linear,I2_linear,hsvImg(:,x_lims(i1,1):x_lims(i1,2),1),I1,I2);
rgbI1I2 = hsv2rgb(hsvI1I2);

Or even simpler if you interpolate over the RGB-planes.

HTH

2 个评论
显示无隐藏无

Felipe Bayona 2020-9-17

编辑：Felipe Bayona 2020-9-17

HiBjorn

Thx for your answer. I realized I didn't make the disclaimer that the 6 images are only examples. Im lloking to generate one four-color gradien matrix.

I'll check to the interpolations you mention.

Bjorn Gustavsson 2020-9-17

Well, if you look at how the Hue, Saturation and Intensity or the Red, Green and Blue varies over your examples, you will start to understand how to design your 2-D colour-maps. I can only assume that there will be quite a bit of tinkering before you manage to generate visually pleasing maps (which are at least two interesting fields of work: what's a visually pleasing map for a given purpose? Does different peoples color-vision/perception vary enough to make big differences in what's good?)

请先登录，再进行评论。