How to convert raw FLIR thermal images(.jpg) into .txt files?

I needed to convert Raw thermal images from FLIR camera which is in .jpg format to be converted to text files, extracting its meta data!

Answer 1

[Ganesh Regoti]

Assuming that you need the meta data of the file into the text format. You can do the following

 k = imfinfo('flir.jpg');  %Get the meta data of the image
 T = NestedStruct2table(k); %Convert the struct to table format
 T = T(1,:); %As the function produces multiple rows of data, pick up only the first row where 
                    %actual data about image resides. 
 FileName = 'test.txt';  %Name of the file in which the meta data must be stored.
 writetable(T,FileName, 'Delimiter', ','); %Write the data into the file placing ’,’ as delimiter

You can refer this to get NestedStruct2table function

https://www.mathworks.com/matlabcentral/fileexchange/51271-nested-structure-to-table-and-or-text-file

In NestedStruct2table function, you need to make small change. In convertCellsToTable function, ignore the case when cell is empty to make it work for your case. Remove the following code

if ~isvector(FieldValuesCells{i})
    error('Field dimension is > 1')
end

[Jubin Mathai]

Thank you Ganesh for the reply. Is there anything missing in the code. I am getting errors while running this code.

[Ganesh Regoti]

That worked for me. Can you give the details of the error? And share the code snippet if possible.

[Jubin Mathai]

The code for converting IR images into text files.

Now the command window reads only the name of the file. My input is definitely wrong. Below is the code I am running. Could you show me a screenshot of how you ran the code?

 k = imfinfo('J.jpg');  %Get the meta data of the image
 function T = NestedStruct2table(k)
 T = T(1,:); %As the function produces multiple rows of data, pick up only the first row where 
                    %actual data about image resides. 
% Extract data from cluster into Cells
[FieldNameCells, FieldValuesCells] = StepThroughCluster(k, [], [], []);
% convert Cells to Table
T = convertCellsToTable(FieldNameCells, FieldValuesCells);
end
%%
function T = convertCellsToTable(FieldNameCells, FieldValuesCells)
%% find max data size
nFields = size(FieldValuesCells,2);% Determine  number of columns 
% Determine max number of rows 
SizeMax = 0;
for i = 1:nFields
    ClassID = class(FieldValuesCells{i});
    switch  ClassID
        case 'double'
            SizeCell = size(FieldValuesCells{i},1);
        case 'char'
            temptest = FieldValuesCells(i);
            SizeCell = size(temptest,1);
        case 'cell'
            SizeCell = max(size(FieldValuesCells{i}));
        otherwise
            warning('class type ''%s'' not implemented', ClassID)
    end
    
    if  SizeCell > SizeMax
        SizeMax = SizeCell;
    end
end
%% Generate  Cell Array
% allocate memory
ValueArray = cell(SizeMax, nFields);
% Fill Cell Array
for i = 1:nFields
    
    ClassID = class(FieldValuesCells{i});
    
    switch ClassID
        case 'double'
            % ValueArray(:,i) = cellstr(num2str(FieldValuesCells{i},10));
            FieldSize = max(size(FieldValuesCells{i}));
            ValueArray(1:FieldSize, i) = num2cell(FieldValuesCells{i});
            
        case 'char'
            ValueArray(1,i) = {sprintf('%s',FieldValuesCells{i})};
            
        case 'cell'
            FieldSize = max(size(FieldValuesCells{i}));
            ValueArray(1:FieldSize, i) = FieldValuesCells{i};
        otherwise
            warning('class type %s not implemented', ClassID)
    end
end
T = cell2table(ValueArray, 'VariableNames', FieldNameCells);
FileName = 'test.txt';  %Name of the file in which the meta data must be stored.
% writetable(T,FileName, 'Delimiter', '\t');
end
%%
function [FieldNameCells, FieldValuesCells] = StepThroughCluster(ClusterIn, FieldNameRoot, FieldNameCells, FieldValuesCells)
FieldNames = fieldnames(ClusterIn);
for ii = 1:length(FieldNames)
    if isstruct(ClusterIn.(FieldNames{ii})) % Step deeper
        fieldnames(ClusterIn.(FieldNames{ii}));
        ClusterOut = ClusterIn.(FieldNames{ii});
        FieldName  =  strcat(FieldNameRoot,FieldNames{ii},'_');
        % use function recursive !
        [FieldNameCells, FieldValuesCells]  = StepThroughCluster(ClusterOut, FieldName, FieldNameCells, FieldValuesCells);
    else % no more Fields => we are at data level => Save data to file
        FieldName  =  strcat(FieldNameRoot,FieldNames{ii});
        FieldNameCells{end+1} = FieldName;
        FieldValues = ClusterIn.(FieldNames{ii});
        FieldValuesCells{end+1} = FieldValues';
    end
end
end

[Ganesh Regoti]

The code must be put outside the function. The edits that are to be made in NestedStruct2table function has already been answered above.

[Stephen23]

@Ganesh Regoti: note that NestedStruct2table.m is copyright code, and must be distributed together with its license. I do not see the licence file attached here, but I presume that this is a temporary oversight that you will fix shortly.

[Ganesh Regoti]

Stephen Cobeldick , Thank you for your catch. I have made appropriate updates.

[Jubin Mathai]

Hi Ganesh,

Could you please share the way you executed the code. I am still getting the same error.