Image Signal Enhancement based on Fractional Differential Technologies.

To describe the ability of image fractional differential in texture detail enhancement and to study the lateral inhibition principle, multiform masks for digital image fractional differential and their operation rules are discussed. The theoretical basis of fractional differential in modulation and demodulation is also analyzed. Through the derivation on the relation between fractional differential and signal time-frequency analysis, we can acquire the separability of two-dimensional fractional differential under certain conditions. Mach's phenomenon generated in image texture detail is studied by us from two aspects: optic nerve model and signal processing, to propose novel masks. Then its operating rules for digital image processing based on fractional differential are proposed. The experiments show that our scheme can effectively reserve better information of edge texture detail during the denoising process, especially for weak texture and gray with little change. The total enhancing effect is obviously superior to integer order differential operator. [ABSTRACT FROM AUTHOR]