基于微分差相关分析和旋转中心定位，提出了用于空间滤波转速遥测中旋转方向的辨识方法。该旋转方向辨识法是将差分空间滤波器中相邻两个光电池的输出信号经微分、相减、取绝对值后再积分计算得到微分差相关值，用其最小值处延时量的符号确定旋转体像在该两光电池处的线速度方向，再依据该线速度方向和旋转中心位置来判定旋转方向。其中的旋转中心位置是利用双差分空间滤波传感器输出信号的两个中心频率及其变化计算系数k，再由k 值和较大中心频率对应的差分空间滤波器所在的位置来确定的。仿真分析实验验证了所提出的旋转方向辨识方法简单易实现。该方法对随机相位不敏感，有效地解决了空间滤波转速遥测中的旋转方向辨识难题。

Based on differential subtraction correlation (DSC) and the position zones of the rotational center, an approach to identifying the rotational direction of the measured body in remote angular-velocity measurement with spatial filters is presented. The DSC is obtained by differentiating and subtracting the output signals of two adjacent photovoltaic cells in one differential spatial filter (DSF), and then integrating the absolute values of the differential subtraction signal. The sign of the temporal delay at the minimum DSC value indicates the displacement direction of the moving image on the photovoltaic cells. The rotational direction is distinguished by using both the displacement direction and the position zone of the rotational center. The position zone of the rotational center is determined by a coefficient k and the location of the DSF with a larger central frequency, where k is calculated with the two central frequencies and their variations of the quasi-sinusoidal output signals of a double-DSF sensor. Experiments and simulations are achieved to confirm the identification method, of which the results show that the identification method is correct and easy to be implemented. The proposed identification method is insensitive to random phase fluctuation, and has solved the difficult problem of identifying the rotational direction in remote angular measurement with spatial filtering technique.