In　this　paper,　we　propose　a　fast　pose　estimate　method　to　estimate　the　relative　pose　between　a　measured　point　cloud　and　a　reference　point　cloud,　in　order　to　increase　the　speed　and　accuracy　of　the　point　cloud　registration　procedure.　The　method　extracts　feature　points　from　the　point　cloud　and　find　the　principal　components　of　the　features　points＇　coordinates.　The　relative　pose　of　the　measured　point　cloud　is　found　by　minimizing　the　principal　components　of　the　measured　point　cloud　and　that　of　the　reference　point　cloud.　After　aligning　the　measured　point　cloud　with　the　reference　point,　we　apply　the　Iterative　Closed　Point　registration　procedure　to　find　the　final　match.　The　method　can　find　the　rotation　and　transformation　relationship　of　two　similar　point　clouds　in　arbitrary　positions　in　a　short　time.　It　can　be　applied　to　determine　the　positions　of　another　similar　target　in　the　case　of　a　given　object　pose