Clear Filters
Clear Filters

Retaining Bit depth when creating a video from Grayscale Images

8 views (last 30 days)
Hello, I am trying to create a video (MP4 file) from a stack of 16 bit Grayscale images. My code is below. I see that using the VideoWriter with the MPEG-4 codec, 16 bit resolution isn't possible and the images need to be down sampled to 8 bit. Is there anyway at all to create a video file and retain the original resolution?
dt = datestr(now,'dd-mmm-yyyy HH-MM-SS'); %get current dateTime
filename=fullfile(filepath,[dt,'- VidMatlab.mp4']);
diskLogger=VideoWriter(filename,'MPEG-4'); % Use the MPEG-4 Codec
ReportMessage(app,'..Disklogger Opened'); %Report message is my own function
ReportMessage(app,['..Recording To Disk Started (', num2str(numFrames),' images)']);
b=0; %Holder if want to add background to change contrast in 8 bit video
for x=1:numFrames
writeVideo(diskLogger, uint8((data{x}+b)/ 256)); %Annoyingly, Need to convert images to 8bit
ReportMessage(app,'DiskLogger Stopped');
disp('Finished Saving Video')
disp('checking file')
v = VideoReader(filename);
fr=v.FrameRate; %double
ReportMessage(app,['Video Saved as: ',filename, ' @',num2str(fr),'fps (',num2str(dur,'%.1f'),' s)']);

Answers (1)

Nandini on 23 Jun 2023
I see that you're using MATLAB to create a video from a stack of 16-bit grayscale images. As you mentioned, the VideoWriter in MATLAB with the MPEG-4 codec does not directly support 16-bit resolution. The images need to be downsampled to 8-bit before creating the video.
In your code, you are converting the images to 8-bit by dividing them by 256 (uint8((data{x}+b)/256)), which effectively scales the pixel values from the original 16-bit range to the 8-bit range (0-255).
Unfortunately, there is no way to create a video file with the MPEG-4 codec and retain the original 16-bit resolution. The MPEG-4 codec is specifically designed for 8-bit color or grayscale images. If you want to preserve the full 16-bit resolution, you would need to use a different codec or file format that supports higher bit depths, such as a lossless format like TIFF or PNG.
Here's an example of how you can modify your code to save the images as a TIFF sequence instead of an MP4 video file:
dt = datestr(now, 'dd-mmm-yyyy HH-MM-SS'); % Get current dateTime
filepath = app.SaveVidEditField.Value;
filename = fullfile(filepath, [dt, '- VidMatlab.tif']); % Save as TIFF file
numFrames = size(data, 2);
ReportMessage(app, ['..Recording To Disk Started (', num2str(numFrames), ' images)']);
for x = 1:numFrames
imwrite(data{x}, filename, 'WriteMode', 'append', 'Compression', 'none');
ReportMessage(app, 'DiskLogger Stopped');
disp('Finished Saving Video');
% Check TIFF sequence properties
info = imfinfo(filename);
numFramesSaved = numel(info);
fr = app.FrameRateEditField.Value;
dur = numFramesSaved / fr;
ReportMessage(app, ['Video Saved as: ', filename, ' @', num2str(fr), 'fps (', num2str(dur, '%.1f'), ' s)']);
In this modified code, each image in the data array is saved as a separate frame in a TIFF file using the imwrite function. The Compression option is set to 'none' to ensure lossless saving. Please note that saving each frame as a separate TIFF image will result in a sequence of individual images rather than a single video file. However, this allows you to retain the original 16-bit resolution of the images.
  1 Comment
Jason on 23 Jun 2023
Thankyou. I actually already have the images as individual tiffs - I then load them into a cell array to make the video!
Is there anyother non MP4 way of retaining the resolution as a video

Sign in to comment.




Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!