肖兴维, 马国鹭, 曾国英. 基于相关因子的无衍射光斑图像定中算法[J]. 应用光学, 2018, 39(4): 500-504. DOI: 10.5768/JAO201839.0402003
引用本文: 肖兴维, 马国鹭, 曾国英. 基于相关因子的无衍射光斑图像定中算法[J]. 应用光学, 2018, 39(4): 500-504. DOI: 10.5768/JAO201839.0402003
Xiao Xingwei, Ma Guolu, Zeng Guoying. Centering method for non-diffracting spot images based on correlation-coefficient[J]. Journal of Applied Optics, 2018, 39(4): 500-504. DOI: 10.5768/JAO201839.0402003
Citation: Xiao Xingwei, Ma Guolu, Zeng Guoying. Centering method for non-diffracting spot images based on correlation-coefficient[J]. Journal of Applied Optics, 2018, 39(4): 500-504. DOI: 10.5768/JAO201839.0402003

基于相关因子的无衍射光斑图像定中算法

Centering method for non-diffracting spot images based on correlation-coefficient

  • 摘要: 为实现无衍射光斑作为直线基准在复杂噪声背景下精确定中, 提出了一种基于相关因子的无衍射光斑图像定中算法。该算法先根据光强重心理论计算光斑中心的大致位置, 再将光斑图像转换成极坐标系下的灰度图, 并生成角频率与光斑图像空间频率相同的离散周期正弦信号, 求解其相位角并对各径向上的相位信息作均方差评价, 计算出极坐标系下理想光斑中心与实际光斑中心的相关因子, 从而达到定中无衍射光斑中心的目的。在模拟噪声环境下, 对比该算法与其他常规算法, 其结果表明:该算法抗背景噪声干扰能力强、计算耗时短, 具有稳定的亚像素定中精度。

     

    Abstract: In order to realize fast and precise centering with non-diffracting spot as straight-line datum under complex background noises, a centering method for non-diffracting images based on correlation-coefficient was proposed. In this algorithm, firstly the approximate position of the beam spot center is calculated based on the theory of light intensity gravity; then the image is converted to the grayscale image in polar coordinates, which generates the discrete-cycle sinusoidal signal with the same angular frequency as the spatial frequency of the spot image, and the phase angle is calculated and the average variance is evaluated for the phase information of each diameter; finally the correlation-coefficient is established in polar coordinate system for ideal beam center and actual beam center, which realizes centering the non-diffracting spot image. In the simulated noise environment, result shows that the proposed algorithm has strong ability of anti-noise interference, less time-consuming in comparison with other common algorithms, which realizes sub-pixel level accuracy.

     

/

返回文章
返回