As one of the most popular optical remote sensor images, MODIS (Moderate Resolution Imaging Spectroradiometer) image are widely used in many areas. However, the processing of MODIS image data is considered as a cumbersome, time-consuming work, especially for the long time series earth observation research. Automatic processing technology is specially needed here. But because of the complex procedure of image matching and the high requirement of location calibration, these images are manual processed in most of the researches. This paper presents an automatic processing method for MODIS image products (mainly for Level 1 B, can be applied on 8-day snow observation image product and daily snow cover optical image data as well). By using the automatic processing system, the efficiency of optical remote sensing image processing is sharply increased while the accuracy in calibration remains the same in comparing with traditional processing method. The working flowchart of the processing system is introduced for those who will deal with mass of MODIS data in their research. Finally, an automatic processing system of snow cover monitoring model based on MODIS L1B image data in ENVI/IDL environment is discussed as the practical application of processing method in long time series snow cover monitoring over Northeast China with MODSI images. The performance shows that the time spent in data processing can be saved from 48 manual working days to 2 working days( 10.41 hours) by computer automatic processing, which proves that processing efficiency of long time series remote sensing data, especial MODIS L1B data, can be greatly increased by saving processing time from months to days and researchers will have more free time from burdensome and automatic work by using the auto processing system.