How to solve (Incorrect use of '=' operator. To assign a value to a variable, use '='. To compare values for equality, use '==')

%-----------------------------------
%           A 3D TV Approach Using Depth Image Based Rendering (DIBR)
%                              Christoph Fehn
%
%      A Survey: Fast & Approximate Algorithm of Depth Image Based Rendering
%                Process for Application of 2D to 3D Conversation
%                       Neetesh Nema, Bragesh Patel
%
%    Approximate Processing of DIBR  Process for 2D to 3D Conversion of Images
%                     Sanjeev Jaiswal, Jigyasha Soni
%
%
%function [V_v] = DIBR(V_o, D_o, K_o, Rt_o, K_v, Rt_v) 
clear all; close all; clc
V_o = imread('V_original.png'); % 2D original image, contains color also known as texture information
% imfinfo ('V_original.png')
D_o = imread('D_original.png'); % corresponding depth map contains normalize depth information
% imfinfo ('D_original.png')
%% Original camera parameters
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Intrinsic parameters of original camera
K_o = [1732.87 0.0 943.23; 0.0 1729.90 548.845040; 0 0 1]; %camera calibrartion matrix
% Extrinsic parameters of original camera
Rt_o = [1.0 0.0 0.0 0; 0.0 1.0 0.0 0.0; 0.0 0.0 1.0 0.0]; % rotation and translation matrix
% depth map normalization factors 
% Znear and Zfar are nearest and fartheset points in the scene from the original camera
Zfar = 2760.510889;
Znear = 34.506386;

%% Virtual camera parameters
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Intrinsic parameters of virtual camera
K_v = [1732.87 0.0 943.23 ;0.0 1729.90 548.845040; 0 0 1];
% Extrinsic parameters of virtual camera
Rt_v = [1.0 0.0 0.0 1.5924; 0.0 1.0 0.0 0.0; 0.0 0.0 1.0 0.0]; % rotation and translation matrix

%% Task 1: produce a synthesized 2D virtual image located at the positioned defined by the two cameras' projection matrices
%% 3D image wrapping
%rgb to ycbcr
% V_o_ycbcr = rgb2ycbcr(V_o); %color space transformation
% D_o_ycbcr = rgb2ycbcr(D_o);

% figure;
% lb={'Y','Cb','Cr'};
% for channel=1:3
% subplot(1,3,channel)
% V_o_ycbcr_C=V_o_ycbcr;
% V_o_ycbcr_C(:,:,setdiff(1:3,channel))=intmax(class(V_o_ycbcr_C))/2;
% imshow(ycbcr2rgb(V_o_ycbcr_C))
% title([lb{channel} ' component'],'FontSize',18);
% end

% figure;
% subplot(1,3,1)
% imshow(V_o_ycbcr(:,:,1))
% title('Y component','FontSize',18);
% subplot(1,3,2)
% imshow(V_o_ycbcr(:,:,2))
% title('Cb component','FontSize',18);
% subplot(1,3,3)
% imshow(V_o_ycbcr(:,:,3))
% title('Cr component','FontSize',18);


% figure;
% lb={'Y','Cb','Cr'};
% for channel=1:3
% subplot(1,3,channel)
% D_o_ycbcr_C=D_o_ycbcr;
% D_o_ycbcr_C(:,:,setdiff(1:3,channel))=intmax(class(D_o_ycbcr_C))/2;
% imshow(ycbcr2rgb(D_o_ycbcr_C))
% title([lb{channel} ' component'],'FontSize',18);
% end


im1 = rgb2gray(V_o);
im2 = rgb2gray(D_o);
% frame_size  = size(im1);
% allpixels = reshape(im1, frame_size);
% frame2 = reshape(allpixels, frame_size);

[H1,W1, ~] = size(im1);
[x1, y1] = meshgrid(1:W1,1:H1);

[H,W, ~] = size(im2);
[x, y] = meshgrid(1:W,1:H);

% m = K_o * eye(3,4) * 
% m1 = K_o * eye(3)*im1;
fx_d = fy_d =1; % assume f = 1, currently unknown
x_k = (- im2./fx_d).*(x1-x);
y_k = (- im2./fy_d).*(y1-y);
XwYwZw = cat(3, x_k,y_k,im2);

%% 2D to 3D transform
XwYwZw = reshape(XwYwZw, H*W, 3);
XwYwZw1 = [XwYwZw ones(H*W,1)];

%% rotate and translate the 3D points
XcYcZc1 = [Rt_o Rt_o(:,4); 0 0 0 1]*XwYwZw1';


%% rotate and translate the 3D points
XcYcZc1 = [Rt_v Rt_v(:,4); 0 0 0 1]*XwYwZw1';

%% part 3 project into image coordinates
%XcYcZc1 = XcYcZc1';
Xc = reshape(XcYcZc1(1,:),H,W); % take Xc and Yc from camera coordinates
Yc = reshape(XcYcZc1(2,:),H,W);
Zc = XcYcZc1(3,:);
Ix = ((x1 .*Xc) ./ im2);
Iy = ((y1 .*Yc) ./ im2);
Dx = Ix+im1;
Dy = Iy+im1;


% m = [x1,y1] + ((eye(3,3)*(Rt_v(:,4)))./2760.510889)+ 

Answer 1

[Walter Roberson]

fx_d = fy_d =1; % assume f = 1, currently unknown

MATLAB does not have "transitive" assignment to set multiple variables to the same value. You will need to use individual assignments (or ugly constructs.)

[Haseeb Hassan]

fx_d =1;
fy_d =1;

I declared and assigned these variables in this way. but now i am getting another error which is given as followeing

----Integers can only be combined with integers of the same class, or scalar doubles.

Error in DIBR (line 92)

x_k = (- im2./fx_d).*(x1-x);-------

The code can be found on this link https://github.com/bipul-mohanto/dibralgorithm/blob/master/DIBR.m

Kindly can you help me to run this code.

Thanks in advance

[Walter Roberson]

The code is broken. It constructs a 3 x 4 matrix and appends the last column of the matrix on again, making a 3 x 5 matrix. Then it tries to put a 1 x 4 matrix under that, which is a size mismatch. If we were to extend the 1 x 4 to be 1 x 5 then the output would be 4 x 5, which would then be * by a 4 x something matrix, which would be an error.

The easiest explanation would be if the extra column should not be appended, which would then give you a 3 x 4 with a 1 x 4 under it making a 4 x 4 * 4 x something. However, the author has been consistent in putting the extra column into several similar constructs, suggesting that there is some Good Reason involved.

[Walter Roberson]

I just opened a few Issues on github against the project. I am not really expecting any response, as the existing issue from 2017 did not receive a response.

I can make some guesses about what the code should be.

[Haseeb Hassan]

Thankyou Sir Walter, i really appreaciate your efforts. Lets wait if someone response to the github query.

Once again many thanks!

[apri zulham]

i have a problem

Error: File: servogui.m Line: 128 Column: 5

Incorrect use of '=' operator. To assign a value to a

variable, use '='. To compare values for equality, use

'=='.

global k;
x=0;

if k=1:1:handles.xSample
   b=readAnalog(3);
   x=[x,b];
   plot(x,'LineWidth',2);
   grip on;
   axis([0 handles.xSample 0 500]);
   pause(0.01);
end

[Walter Roberson]

Are you trying to make a single comparison? If so then you would use == -- but notice that the right hand side of the = is a vector so there is no chance that the test can succeed unless k is a vector the same size (and even that is unlikely to work)

The form of your code is as if you are expecting to loop. Loops use for not if