As FreeBeerReekingMonk suggests, the %seen style of hash is often used for counting duplicates:

my %seen;
$seen{$_}++ for values %hash;
my @dups = grep { $seen{ $hash{$_} } > 1 } keys %hash;
print join(',', sort @dups );
[download]