Abstract: Aimming at the problem of multi-target recognition with many characteristic indexes,a new fusion method for the fuzzy muti-sensor data is proposed according to the entropy weight.The method defines the entropy weight according to the theory of entropy,introduces the notations of the optimal and suboptimal membership degrees and uses the relative approach degree to measure the distance.Hence the rule of target recognition is given.The method may avoid the subjectivity of the weight of characteristic indexes and improve the objectivity and accuracy of target recognition.The example of parts recognition proves that the method is both effective and exercisable.