The Classification and Recognition of Wax Crystal Structure through Image Analysis

Based on fractal theory, a new method was proposed for extracting wax crystal image features. The wax crystals were separated out from the transported crude when temperature went down. The wax crystals have self-affined and self-similar characters when they are imaged through a polar-microscope. The crystallizing process is also affected by some environmental factors such as temperature and shearing function when the oil pumped through a pipeline. Fractal attributes of the wax crystal image are computed from the morphologic dilation and used for classification. The wavelet coefficient decomposition and gray-level co-occurrence matrix features were also discussed in this paper