This paper presents a method to describe and extract the general feature structures based on the fuzzy theory. This paper first builds some fuzzy measure rules with three levels (the basic property Φ, the derived property γ , and the associated property Π ). With the aid of measure rules, the feature vectors are established for fuzzy membership calculation and the feature region extraction. The analysis demonstrates that this method can obtain the optimal division of the flows on the principle of the minimum square sum. Further, the experiments show that the extent of typical flow structures extracted by this method are more effective than existing methods. In addition, the transfer function can be more flexible in design to avoid the cluttering and occlusion problems, which must be solved when visualizing 3D flows.