I did not use any specific software for that. Only blast against Refseq or uniref was enough. Many mitochondrial genes are highly conserved and could be found easy. Usually they are not spread in the genome, by chromosome walking you may find their neighbors. I founded some complications to determine introns as in some cases they hold signatures or domains (in fungi).
the first step in NGS studies is to align your data to a reference genome. in your case you'll be able to align against the mtDNA and filter the rest of data. follow this wiki for more explanations (https://en.wikibooks.org/wiki/Next_Generation_Sequencing_(NGS)/Alignment).
once I just took the mitochondrial sequence from Arabidopsis and blasted it against the barley genome to extract the mitochondrial DNA from that. It worked very well.