Method　of　robotic　hand-eye　calibration　based　on　active　vision　is　presented　in　this　paper.　Firstly　the　calibration　method　of　the　hand-eye　relations　rotation　matrix　is　analyzed.　The　world　coordinates　of　feature　points　on　the　workpiece　platform　are　obtained　by　laser　pointer,　that　is　mounted　on　the　mechanical　arm.　Then　we　can　calibrate　the　hand-eye　relations　translation　vector.　Through　the　calibrated　hand-eye　relations,　we　can　calculate　the　distances　between　four　feature　points　in　the　image,　and　compare　the　error　with　the　actual　values.　Experimental　results　show　that　the　measurement　error　range　of　the　lengths　between　the　feature　points　is　±0.8　mm.　Therefore,　the　hand-eye　calibration　precision　is　high,　and　it　can　meet　the　requirements　of　workpiece　positioning　and　automatic　grabbing　for　robot.　©　2015　IEEE.