Shape classification of snow particle into snowflake and graupel using image processing

Classification of snowflakes is important in the investigation of the growth process of snow particles and radar meteorology, etc. This paper proposes an automatic image processing method for classification of falling snow particles. A side view of natural falling snow particles was photographed by IEEE1394 camera (1280 times 960 pixel) with a shutter speed of 1/10000 s. To obtain shape parameters, both a binary and a gray-level image of each particle were recorded continuously. Each particle was classified into a graupel or a snowflake by statistical classification procedure using image features. The proposed method has been successfully tested in snowfall, obtaining the percentages of snowflakes and graupels at 1 min interval.