3D Discrete Spherical Fourier Descriptors Based on Surface Curvature Voxels for Pollen Classification

This paper presents a new method to extract 3D Discrete Spherical Fourier Descriptors (DSFD) based on surface curvature voxels for pollen recognition. In order to reduce the high amount of pollen information and noise disturbance, the geometric normalized curvature voxels with the principal curvedness are firstly extracted to represent the intrinsic pollen volumetric data. Then the curvature voxels are decomposed into radial and angular components with Spherical Harmonic Transform in spherical coordinates. Finally the discrete 3D Fourier transform is applied on the decomposed curvature voxels to obtain the 3D Spherical Fourier Descriptors for pollen recognition. Experimental results show that the presented descriptors are invariant to pollen image rotation, scale and translation, and can bring about good recognition precision and speed simultaneously.