用圆谐综合滤波器实现三重不变光学图像识别

Multiple invariant optical pattern recognition using circular harmonic synthetic filter

本文在圆谐滤波器算法基础上,结合综合判别函数提出圆谐综合滤波器,解决传统匹配空间滤波器对输入目标的几何形变问题。对合成的滤波器进行了计算机模拟和光学相关实验,结果表明圆谐综合滤波器具有较强的三重不变光学图像识别能力。

A novel filter, circular harmonic synthetic filter, is presented to overcome the sensitivity of conventional matched spatial filter to geometrical dsitortions of the input object. The computer simulation and optical correlation experiments were made for the synthetic filter, and the results showed that the synthetic circular harmonic filter has stronger recognition capability for triple invariant optical patterns.