unable to run the label overlay of original and labeled image for data training on matlab

i want to train the data using vgg16,which expects input in rgb.

first i changed the graylevel mri images to rgb using the command ,rgbImage = ind2rgb(x, gray(256));

then i labeled the image using image labler on matlab,then the pixellabeldata is stored in png format,and i don't know why but the labeled image adjusted itself in graylevel and png format,when i tried to convert the pixellabeldata in to rgb using rgbImage = ind2rgb(x, gray(256));,command but the labeloverlay returns error saying.Error in labeloverlay>parseInputs (line 150)

validateattributes(Ldouble,allowedTypes,{'integer','nonsparse','real','nonnegative','nonempty','ndims',2},mfilename);

Error in labeloverlay (line 88)

parsedInputs = parseInputs(varargin{:});

Error in viewmask (line 23)

labeloverlayImg = labeloverlay(img, label

what is the problem is that because try to convert labeled image to rgb?can i simply apply the rgb image and the pixellabel image which adjusted itself to png and graylevel format for inputing on vgg16 or what is the solution

Answer 1

[Image Analyst]

I'm not sure what your x is, but something like this works fine:

x = imread('moon.tif'); % x is a gray scale image.
rgbImage = ind2rgb(x, gray(256));

Alternatively you can do this:

rgbImage = cat(3, x, x, x);

[eten hyle]

i already tried it ,i apply the above command the labeled image converted to rgb but when i run the labeloverlay(labeloverlayImg = labeloverlay(img, label); it returns error

Error in labeloverlay>parseInputs (line 150)

validateattributes(Ldouble,allowedTypes,{'integer','nonsparse','real','nonnegative','nonempty','ndims',2},mfilename);

Error in labeloverlay (line 88)

parsedInputs = parseInputs(varargin{:});

Error in viewmask (line 23)

labeloverlayImg = labeloverlay(img, label

[Image Analyst]

This does not look like the correct syntax:

labeloverlay(labeloverlayImg = labeloverlay(img, label);

Check it.