A new approach to extract the chaotic characteristics of the two-phase stirring and mixing state is proposed for bottom-blown oxygen-enriched bath smelting process of copper. By quantifying the local mixing characteristics in the stirred reactor of bottom-blowing copper smelting, an improved 0-1 chaotic test method was introduced to measure the chaotic characteristics of a time series of mixing index. It was found that the different channels of the RGB image of turbulence flow field are not the same for the contour feature extraction of the bubble; the single-channel horizontal profile of the single-and doubledistributor flow field images shows single and double peaks, respectively, verifying the accuracy of the hybrid characterization. After calculating the mean and standard deviation time series of grey intensity of the region of interest, the median of the asymptotic growth rate K corr (c) of the mixing index time series was used as the criteria of chaos detection in the molten pool dynamic balance state. The variability of chaos in different mixing processes has been more accurately characterized.