image processing, dimension reduction

3 次查看（过去 30 天）
mohammed abdul wadood
评论： mohammed abdul wadood 2018-3-20
if i have number of images (multi spectral images for the same scene), and i want to reduce the dimensional of this images by any dimension reduction techniques where in the result i get one reducing image for 200 as example. how can i do this task? thank you.

请先登录，再回答此问题。

采纳的回答

Image Analyst
Image Analyst 2018-3-19
See my attached demo for pca. It should be easy for you to adapt it to take any number of spectral bands instead of the 3 visible color (RGB) bands I took. If you want just one single image that represents most of the variation, take the principal components #1 image.
  4 个评论
显示 2更早的评论
mohammed abdul wadood
mohammed abdul wadood 2018-3-20
ok, thanks all; another question please, i wrote an code for (KLT) techniques for dimension reduction for image (I'm just a beginner in matlab my code maybe is not good enough one or it's want some modification) the result for the code is binary image, is it possible that i reduce the dimensionality without losing the color (can i get colored reducing image as result)
clc, clear all, close all x = imread ('10.bmp'); whos %x2 = imread ('2.bmp'); %x3 = imread ('3.bmp'); %x4 = imread ('4.bmp'); %x5 = imread ('5.bmp'); %x6 = imread ('6.bmp'); %x=[x1 x2 x3 x4 x5 x6]; t=graythresh(x) z=im2bw(x,t); [n m]=size(z);
for j=1:m mu(:,j)=mean(z(:,j)); end for j=1:m xx(:,j)=z(:,j)-mu(j); end xx=xx'; cx=(xx*xx')./(n-1); [M V]=eig(cx);
%my
V; size(V) mm=max(max(V)) ee=sum(diag(V));
%
A=(rot90(M))'; y=A*xx; Ay=A'*y; for j=1:m xhad=Ay+mu(j); end
k=input('k='); for i=1:k for j=1:k xxk(i,j)=xx(i,j); cxk(i,j)=cx(i,j); end end for j=1:k muk(:,k)=mean(xxk(:,k)); end
[M1 V1]=eig(cxk); Ak=(rot90(M1))'; yk=Ak*xxk; Ayk=Ak'*yk; for j=1:k xhadk=Ayk+mu(j); end
figure; imshow(xhadk),title('A=510')
figure; subplot(3,2,1),imshow(x); subplot(3,2,2),imshow(xhad); subplot(3,2,3),imshow(xhadk);

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Dimensionality Reduction and Feature Extraction 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by