Unable to read FITS file in Matlab but ImageJ(Fiji) is able to read it

My Fits File1: Link

My Fits File2: Link

Debug

 fitsinfo('surfaceFull.fits')

Output

Output argument "precision" (and possibly others) not assigned a value in the execution with "fitsinfo>bitdepth2precision"
function.

Error in fitsinfo>updateInfoForPrimary (line 353)
            info(1).DataType =  bitdepth2precision(value);

Error in fitsinfo>readheaderunit (line 291)
        info = updateInfoForPrimary(info);

Error in fitsinfo (line 220)
info.PrimaryData = readheaderunit(fid, info.FileSize);

My Code

clear all; close all; clc;
%% Read and display a FITS file in MATLAB
filePath = 'data\getDepthFromSparse3Doct\surfaceFull.fits'; 
data = fitsread(filePath);
figure;
imagesc(data); % This function scales the display based on the range of data values
colorbar; % Adds a colorbar to the figure
title('Display of FITS File Data');
axis image; % Sets the aspect ratio so that data units are the same in every direction

Output

Output argument "precision" (and possibly others) not assigned a value in the execution with "fitsinfo>bitdepth2precision"
function.
Error in fitsinfo>updateInfoForPrimary (line 353)
           info(1).DataType =  bitdepth2precision(value);
Error in fitsinfo>readheaderunit (line 291)
        info = updateInfoForPrimary(info);
Error in fitsinfo (line 220)
info.PrimaryData = readheaderunit(fid, info.FileSize);
Error in fitsread (line 103)
    info = fitsinfo(filename);
Error in depthMapReconstrcution (line 4)
data = fitsread(filePath)

Answer 1

[Sulaymon Eshkabilov]

Just looking at the ImageJ imported images from the two FITS files, two points can be made:

(1) FITS image data file is not complete

image(imread('surfaceFull.png'))

(2) FITS image data is in 32-bit, and hence, bit size precision issue is prompted.

Answer 2

[Walter Roberson]

The first fits file (at least) consistently is empty in columns 9 and 10, and then starts columns 11 and 12 with '= ' . That appears at the moment to be acceptable to the definitions at https://archive.stsci.edu/fits/fits_standard/node29.html#SECTION00912200000000000000 but on the other hand https://archive.stsci.edu/fits/fits_standard/node39.html#SECTION00941120000000000000 says that for BITPIX that the value field must contain an integer...

Ummmm, let me think...

I suspect there might possibly be no contradiction in the definitions if columns 9 and 10 are empty but columns 11 and 12 do not start with '= ' -- if the value appears directly.

In any case the code in fitsinfo function getValueAndComment is treating the rest of the BITPIX line as a comment. Regardless of whether that is right or wrong, that is the difficulty of the file: that in the primary header, columns 9 and 10 are expected to be '= ' and columns 11 onwards are expected to be the numeric value, whereas in the actual file, columns 9 and 10 are blank and columns 11 and 12 contain '= ' and then there is a value after that.