My code is showing error " matrix dimension must agree".In line no -4 (2nd function:function [res_img]=​Gray_trans​_S_curve_u​int16(xyz)​)-nrm_val= (xyz)- mn1;.output of the 1st function will be used in 2nd function.1st fn :function Res=Local_Process_uin

function Res=Local_Process_uint16_mymod(inputImage,nblockrow,nblockcolumn) [m,n] = size(inputImage); dcol = fix(n/nblockcolumn); drow = fix(m/nblockrow); for index = 1:nblockrow* nblockcolumn

    [r,c] = ind2sub([nblockrow,nblockcolumn],indexa );
    if (r==nblockrow && c~=nblockcolumn)
        subimage = inputImage((r-1)*drow+1:r*drow+rem(m,nblockrow), (c-1)*dcol+1:c*dcol,:);
    elseif (r~=nblockrow && c==nblockcolumn)   
        subimage = inputImage((r-1)*drow+1:r*drow, (c-1)*dcol+1:c*dcol+rem(n,nblockcolumn),:);
    elseif (r==nblockrow && c==nblockcolumn)
        subimage = inputImage((r-1)*drow+1:r*drow+rem(m,nblockrow), (c-1)*dcol+1:c*dcol+rem(n,nblockcolumn),:);
    else
        subimage = inputImage((r-1)*drow+1:r*drow, (c-1)*dcol+1:c*dcol,:);
    end;
    subimage;
    test_th=Gray_trans_S_curve_uint16(double(subimage));
    Box_new{r,c}=test_th;
    subimage=[];

end

Res=cell2mat(Box_new); % figure,imshow(Res,[]); % 2nd function

function [res_img]=Gray_trans_S_curve_uint16(xyz) mn1 = min(xyz(:)); mx1=max(xyz(:)); mxl nrm_val= (xyz)- mn1; diff = (mx1-mn1); res=[]; res_img=[]; xyz_norm = nrm_val ./ diff;

[m,n]=size(xyz_norm);

for i=1:m for j=1:n res(i,j)=(0.9642)+((8.594*10^-4-0.9642)/(1+exp((xyz_norm(i,j)-0.4969)/0.07598))); end; end;

res_img=uint16(((mx1-mn1)*res)+mn1);

%endThe attached file contains contrast enhancement code. function Res=Local_Process_uint16_mymod(inputImage,nblockrow,nblockcolumn) will take input as image and size of blocks of subimage.Output of the image will be sed in another function name:function [res_img]=Gray_trans_S_curve_uint16(xyz). but it is showing error. example:function Res=Local_Process_uint16_mymod('mrimage.bmp',40,40) objective: The function function Res=Local_Process_uint16_mymod(inputImage,nblockrow,nblockcolumn) will take input image and divide the image into non overlapping blocks. after that each block will be given as input to the function [res_img]=Gray_trans_S_curve_uint16(xyz) to perform contrast enhancement. but in line no 4 it is showing error Marix dimenson must agree.%

[KALYAN ACHARJYA]

Share the main code where you call these two function, it would be helpful to get answered.

[Milan Singh]

these are the code

[Majid Farzaneh]

Hi, do you have a file 'mxl.m'?

I have an error in Gray_trans_S_curve_uint16.m- Line5. mxl is not defined. By removing line5, it works. But I don't know it works correct or not.

[Majid Farzaneh]

َAnother problem is that you didn't return Res value in the function 'Local_Process_uint16_mymod' You should assign output to Res variable.

[Milan Singh]

@Majid Farzaneh I have send another file. Please tell me what input you are taikng in command window.

Answer 1

[Majid Farzaneh]

If your input is an image it's incorrect to use it like this:

Res=Local_Process_uint16_mymod('mrimage.bmp',40,40)

'mrimage.bmp' is a string not an image. You should first read the image:

I=imread('mrimage.bmp');
Res=Local_Process_uint16_mymod(I,40,40)

[Milan Singh]

Sir,this time i amgetting different error Error in Local_Process_uint16_mymod (line 18) subimage = inputImage((r-1)*drow+1:r*drow, (c-1)*dcol+1:c*dcol,:);

[Majid Farzaneh]

I don't have any error by this:

Probably your image is the problem. Can you send me 327.bmp or mriimage.bmp??

[Majid Farzaneh]

Hi again. The problem is your input image size. Your input has 3 dimensions. But your function works for 2-D image. Before using the function use rgb2gray or mat2gray function for your original image:

I=imread('327.bmp');
Ig=rgb2gray(I);
figure, imshow(Ig), title('original image');
Res=Local_Process_uint16_mymod(Ig,40,40);
figure, imshow(mat2gray(Res)), title('Enhanced image');

[Majid Farzaneh]

[Majid Farzaneh]

By the way you can also use histeq() and adapthisteq() functions for contrast enhancement. note that the input must be 2-D.

Good luck!

[Milan Singh]

thank you Majid_