Laser-induced breakdown spectroscopy(LIBS) has been studied for the detection of Cu particles in transformer oil, but its practical application was limited by the poor limit of detection. To improve the detection capability of LIBS, research on the time and space resolution of plasma would help optimize the LIBS measurement parameters. In the research, transformer oil with Cu particle contamination was titrated to filter paper for sample preparation, and the distribution of Cu, N atoms and CN, C2 molecular bands in the plasma evolution were studied combining the plasma spectral and imaging analysis method. The experimental results shown that the C2 molecular band is mainly produced by the pyrolysis of transformer oil, and CN is mainly produced by the reaction of C2 and N2; Cu atoms are mainly distributed within 0.5-3mm upon the sample surface, and CN molecular band spectrum has high stability in the region. Due to the optimization of experimental parameters, the application prospect of LIBS as a particle concentration detection method for transformer oil has become practical. Furthermore, the detection of different particulate pollutants will help determine the operating status of the transformer.