A Tracking Algorithm Based on Fractional Differential Gradient within 3D Image

摘要

Based on fractional derivative gradient, tracking operator is proposed to extract three-dimensional boundary surface within biomedical image. Firstly, the gradient and zero-crossing are calculated via the integer order derivative to extract the main structure of three-dimensional boundary surface. Then the slice images are enhanced by fractional differential, based on coplanar principle, the detailed subtle structures are tracked in three-dimensional images. The experimental result indicates that the method can track more detailed subtle three-dimensional structures. The improved tracking algorithm illustrates a good prospect of application and extension to other 3D biomedical image analyses and researches by comparing with the traditional boundary surface tracking algorithm.