With the development of remote sensing space resolution, the atmosphere as impassable medium of remote sensing imaging made the resolution decrease. For quantitative analysis on the influence of different atmospheric conditions, the Monte-Carlo theory and atmospheric radiative transfer by Monte-Carlo method was introduced firstly, and the photons diffusion in different conditions was simulated. Atmospheric point spread function is studied under different conditions of wavelength, different visibility, different remote sensor height, different asymmetry indices of aerosol particle and different observation angles based on standard atmospheric models of Modtran. Experimental results indicate that different atmospheric conditions have some influence on data quality of remote sensing image.