Self-excited dust scrubbers are widely used in the purification of a high dust concentration. However, the existing research has been primarily focused on the macroscopic performance of the dust scrubbers and has neglected the gas-liquid two-phase flow patterns that are closely related to its dust collection principle. This paper employed high-speed photography technology to acquire video images of liquid phase patterns and employed high-frequency acquisition technology to acquire the corresponding dynamic pressure signals of the gas-phase of the self-excited dust scrubber. Using the probability density distribution (PDD), wavelet analysis, and power spectral density (PSD) estimation, in combination with the dynamic image analysis of the gas-liquid two-phase flow, a quantitative study of the gas-liquid two-phase flow patterns of self-excited dust scrubbers was conducted. This study provides strong supporting data to overcome the bottleneck that limits theoretical studies on self-excited dust scrubbers. (C) 2016 Elsevier Ltd. All rights reserved.