Error in displaying .TIF images in MATLAB

I tried to display a .tif image in MATLAB. The system is able to read it, but not able to display it.

pet=imread('13.tif');
imshow(pet);

Error using imageDisplayValidateParams>validateCData (line 113)
Multi-plane image inputs must be RGB images of size MxNx3.

Error in imageDisplayValidateParams (line 27)
common_args.CData = validateCData(common_args.CData,image_type);

Error in imageDisplayParseInputs (line 78)
common_args = imageDisplayValidateParams(common_args);

Error in imshow (line 219)
  [common_args,specific_args] = ...

[Eric Auth]

Hi Gee,

Were you ever able to solve this problem? I am getting a similar error when trying to view a .TIF file. I have been able to read .TIFF images that were converted from RAW images, however when I use adobe to convert screen captures from my camera's video mode, Matlab does not seem to like the image. Let me know. Thanks

[Walter Roberson]

Are the TIFF written as CMYK or as RGBA or are they multispectral?

[Eric Auth]

How would I check this? The .TIFF files are .ARW (raw) files from a Sony camera converted to .TIFF by the program dcraw. The .TIF are video screenshots from Adobe PremierePro.

[Walter Roberson]

What are the options you are using for dcraw ?

[Eric Auth]

"-6: Write 16-bit instead of 8-bit" and "-T: Write TIFF instead of PPM"

[Walter Roberson]

I think I would need to see the TIFF metadata. As a first pass could I see the output of imfinfo() applied to the tiff ?

[Eric Auth]

Filename: 'C:\Users\lenovo\Documents\MATLAB\BB_0.1.tiff'

                  FileModDate: '27-Nov-2017 22:08:51'
                     FileSize: 60545788
                       Format: 'tif'
                FormatVersion: []
                        Width: 5496
                       Height: 3672
                     BitDepth: 24
                    ColorType: 'truecolor'
              FormatSignature: [73 73 42 0]
                    ByteOrder: 'little-endian'
               NewSubFileType: 0
                BitsPerSample: [8 8 8]
                  Compression: 'Uncompressed'
    PhotometricInterpretation: 'RGB'
                 StripOffsets: 1852
              SamplesPerPixel: 3
                 RowsPerStrip: 3672
              StripByteCounts: 60543936
                  XResolution: 300
                  YResolution: 300
               ResolutionUnit: 'Inch'
                     Colormap: []
          PlanarConfiguration: 'Chunky'
                    TileWidth: []
                   TileLength: []
                  TileOffsets: []
               TileByteCounts: []
                  Orientation: 1
                    FillOrder: 1
             GrayResponseUnit: 0.01
               MaxSampleValue: [255 255 255]
               MinSampleValue: [0 0 0]
                 Thresholding: 1
                       Offset: 10
             ImageDescription: '                               '
                         Make: 'Sony'
                        Model: 'DSC-RX10M3'
                     Software: 'dcraw v9.27'
                     DateTime: '2017:11:27 18:43:07'
                       Artist: ''
                DigitalCamera: [1×1 struct]
             ICCProfileOffset: 1376

[Eric Auth]

By comparison, what we have for the .TIF from the video/AdobePremierePro:

Filename: 'C:\Users\lenovo\Documents\MATLAB\Chopstick.tif'

                  FileModDate: '18-Feb-2018 21:58:24'
                     FileSize: 8307822
                       Format: 'tif'
                FormatVersion: []
                        Width: 1920
                       Height: 1080
                     BitDepth: 32
                    ColorType: 'truecolor'
              FormatSignature: [73 73 42 0]
                    ByteOrder: 'little-endian'
               NewSubFileType: 0
                BitsPerSample: [8 8 8 8]
                  Compression: 'Uncompressed'
    PhotometricInterpretation: 'RGB'
                 StripOffsets: [1×1080 double]
              SamplesPerPixel: 4
                 RowsPerStrip: 1
              StripByteCounts: [1×1080 double]
                  XResolution: []
                  YResolution: []
               ResolutionUnit: 'Inch'
                     Colormap: []
          PlanarConfiguration: 'Chunky'
                    TileWidth: []
                   TileLength: []
                  TileOffsets: []
               TileByteCounts: []
                  Orientation: 1
                    FillOrder: 1
             GrayResponseUnit: 0.01
               MaxSampleValue: [255 255 255 255]
               MinSampleValue: [0 0 0 0]
                 Thresholding: 1
                       Offset: 8303182
                     DateTime: '2018:02:18 16:58:24'
                          XMP: '<?xpacket begin="" id="W5M0MpCehiHzreSzNTczkc9d"?>…'
                    Photoshop: [56 66 73 77 4 37 0 0 0 0 0 16 212 29 140 217 143 0 178 4 233 128 9 152 236 248 66 126]

[Walter Roberson]

I am confused. The headers for BB_0.1.tiff are those for an RGB image with 8 bits per channel, 5496 wide and 3672 high. I see no evidence in the headers for it that it would return 4 planes. This file was written by dcraw.

On the other hand the headers for Chopstick.tif show four channels of 8 bits per channel, 1920 wide and 1080 high. It is not immediately clear to me what the 4th channel is, but alpha would make the most sense.

[Eric Auth]

Thanks a lot! By opening the screenshot from premeirepro in Adobe Photoshop, you are able to save as a .TIFF and there is an option to "remove alpha channels". After doing this I am able to open the screenshot in Matlab.

[Walter Roberson]

I suspect that if you were to use imread() and discard the 4th plane then that would be RGB.

[Eric Auth]

Is there an easy way to do that? That would make the process much more efficient

[Walter Roberson]

tiff_image = imread('AppropriateFileName.tif');
if ndims(tiff_image) > 3
  error('got tiff image with more than 3 dimensions???');
end
if size(tiff_image,3) == 4
  tiff_image = tiff_image(:,:,1:3);
elseif size(tiff_image,3) == 1
  tiff_image = tiff_image(:,:,[1 1 1]);  %grayscale to rgb
elseif size(tiff_image,3) ~= 3
  error('expected tiff image to have 1 channel, or 3 channels, or 4 channels, got %d\n', size(tiff_image,3));
end

[Walter Roberson]

Note that this can be made much simpler if you assume that all of the data is the same:

 tiff_image = imread('AppropriateFileName.tif');
 tiff_image = tiff_image(:,:,1:3);

[Soumyadip Ray Cahudhuri]

this is my code

I=imread('pout.tif');
imshow(I)
figure
imshow(I>220)

this is thr error

Error using images.internal.imageDisplayValidateParams>validateCData (line 115)

Multi-plane image inputs must be RGB images of size MxNx3.

Error in images.internal.imageDisplayValidateParams (line 27)

common_args.CData = validateCData(common_args.CData,image_type);

Error in images.internal.imageDisplayParseInputs (line 78)

common_args = images.internal.imageDisplayValidateParams(common_args);

Error in imshow (line 223)

[common_args,specific_args] = ...

Error in xd (line 2)

imshow(I)

what is the solution ?

[Walter Roberson]

Show us size(I) and imfinfo('pout.tif')

Answer 1

[Walter Roberson]

What does size(pet) report? If it shows a 4 dimensional image, and you have the Image Processing Toolkit, try implay(pet)

[Gee Cheng Mun]

The size reported is [ 235 106 4 ]. I need to perform color detection on this image, and extract out the red, yellow and green color bands. I tried to use implay but it came out as a black and white image.

[Ronnie II Concepcion]

Hi! I am trying to extract RGB channels from a colored picture and the resulting size and data type of each RGB component is MxNx7 uint8. I need to change the size to MxNx3. I think we have the same challenge. May I know how you resolved it? Thanks.

[Walter Roberson]

Please show imfinfo() of the file.

Possibly the code I showed above would work,

tiff_image = imread('AppropriateFileName.tif');
tiff_image = tiff_image(:,:,1:3);

However, in my experience, TIFF with 7 channels are typically multispectra and do not necessarily have RGB or the RGB might not be the first three channels. I have found that most people who have TIFF with 7 channels want to synthesize an RGB image from multispectra data, typically hoping to include two channels of infrared from satellites.