How can I handle variable matrix size error in HDL code generation?

I am working on a image processing algorithm. I want to convert that into Verilog HDL through MATLAB HDL coder. But the error I am facing while converting is Error '' : Error: variable-size matrix type is not supported for HDL code generation. I also tried to define the size of the variable beforehand using zeros(); but this is also not working. Please help me to get over this issue.

function  img_or = Intra_Order(image_wide, image_high, org)

img_or = zeros(1,1966080);
img_or = uint8(img_or);

org = reshape(org,[1080,1920]);
img = reshape(org',[1,1920*1080]);

write_addr = 1;

%file_size = image_wide * image_high;

% Order input frame
for i=0:(image_high/64 - 1)
    for j = 0:(image_wide/64 - 1)
        for k = 0:63

              read_addr = j*64 + i*image_wide*64 + image_wide*k + 1;
        img_or(write_addr:write_addr+63) = img(read_addr:read_addr+63);
        write_addr = write_addr + 64;

          end
      end
  end

Answer 1

[Tim McBrayer]

You are using image_wide and image_high to determine your loop limits, which are input variables. You will at least have an issue with your loops being reported as unbounded. You are aiming for hardware; everything needs to have a size.

[Vishal Deep]

I changed image_wide and image_high to 1920 and 1080 respectively. But still the same error. The error is in this line or in the variable img_or

img_or(write_addr:write_addr+63) = img(read_addr:read_addr+63)