Usually, flow patterns of solid-liquid two-phase flow are discriminated by means of observing particle-bed situations, namely judging whether stationary or sliding bed. The usual method depends on observer's subjectivity. This work presents a new and objective method based on fluctuating intensity of particle velocity vectors measured by a particle tracking velocimetry (PTV). The relationship between the fluctuating intensity of deflection angles of the vectors and the modified Froude number is a good index to discriminate the flow patterns objectively. The flow patterns have been defined by crisp sets, however the definitions contain several contradictions. Analyzing uncertainty and ambiguity which are involved in the flow patterns and the process of judgment, the new definitions by means of considering fuzzy sets are proposed.