How can i makes matching between features extracted from pictures

Hello everybody

I am working on fingerprint identification project, I use Hu's Seven Moments Invariant algorithm, to extract the features, after that I save this features as a matrix for each picture, the database containe 60 cells, in every cell there is a matrix of 7 rows and one column, I need your help to compute the similarity between new matrix and the database, and retern the result depends on threshold.

at this time I use this code to compare between test template and train template, but it give me only the minimum distance, where I can't specify threshod.

 euclidean_template=cellfun(@norm,trdata ) ;       
  test_template=cellfun(@norm,tedata )   ;
for i=1:length(test_template)
    [minDistance, indexOfMin] = min(abs(euclidean_template-test_template(i)));
    formatSpec = '%4.2f is  %8.3f \n';
    fprintf(formatSpec,i,indexOfMin)  
end

thank you