Analysis of Time Evolution of Patterns Based on Various Image Processing Techinques

Images of patterns are processed with some image processing technology, such as spatial intensity distribution, spatial correlation function and Fourier transformation, thus some important intrinsic elements of active plasma and pattern formation can be presented more clearly. Images of disordered grids, spiral pattern and big spot hexagonal pattern are obtained in dielectric barrier discharge system at different times under the same experiment condition. Using spatial intensity distribution and spatial correlation function, it is found that the average distance becomes unique and the structure turns more regular with time relaxed, and most hexagonal cells in big spot hexagonal pattern are relatively perfect. For further investigation, these images are transformed to frequency domain and Fourier spectra are studied. It is found that the mode changes with time relaxed. Grids pattern presents disordered structure. The spiral pattern is formed with single-wavelength but orientation unselected. The stable big spot hexagonal pattern is formed with single-wavelength three-wave resonance. With time relaxed, separation of system modes leads to increase of system periodicity by means of self organization.