With　the　development　of　Intelligent　Transportation　Systems,　it　is　more　important　to　improve　the　calculation　accuracy　of　traffic　flow,　speed　and　occupancy　through　traffic　surveillance　videos.　The　biggest　challenge　to　calculate　these　parameters　is　getting　accurate　results　in　congested　traffic　situations.　We　have　proposed　a　method　to　solve　this　problem,　which　consists　of　three　parts.　First　of　all,　a　dynamic　background　subtraction　method　is　used.　Secondly,　we　use　blob　tracking　and　module　cluster　or　splitting　to　imitate　vehicle　contour　in　target　tracking.　Finally,　we　use　improved　three-frame　difference　method　to　calculate　vehicle　speed.　Experimental　results　show　that　the　proposed　methods　are　superior　to　other　traditional　methods,　especially　in　congested　traffic　situations.　©　2016　IEEE.