In order to identify the flow regime of gasliquid twophase flow effectively, air and water are taken as the experiment media and a vortex flowmeter is employed as component to induce the flow around bluffbody. The ductwall differential pressure method is adopted to acquire the wake fluctuation signals of the flow around bluffbody in the gasliquid twophase flow. The signals are decomposed with the ensemble empirical mode decomposition, and then the Hilbert marginal spectrum is obtained with Hilbert transform. The maximum crosscorrelation coefficient method is used to sift the characteristic intrinsic mode functions. The energy ratios of the characteristic intrinsic mode functions are selected, and used to construct the flowpattern maps together with the volumetric void fraction and the twophase Reynolds number respectively. The results show that the two newly constructed flowpattern maps can successfully identify the typical flow patterns such as singlephase water, bubble flow, plug flow and slug flow with the recognition rate up to 91.67% and 88.89% respectively, which can meet the needs of most practical engineering applications nicely.