索引圆环灰度差值最小的光斑中心逼近法

Spot center approximation method with minimum gray-scale difference of indexed ring

  • 摘要: 为了提高光栅衍射型激光告警系统参考位置的定位精度,提出了一种基于索引圆环灰度差值最小的光斑中心逼近法。根据给定光斑中心的初始坐标生成3×3中心坐标矩阵,使用Lagrange插值法求出以中心坐标矩阵为中心的索引圆环上各点的灰度值,通过二分法逐步逼近索引圆环灰度差值最小的中心坐标,即光斑中心真实坐标。在准确性对比实验中,本文所提出的算法在3组理想光斑模型样本中x方向、y方向平均误差分别为0.0013像素、0.0011像素。在稳定性对比实验中,本文算法在3组实际光斑图片样本中x方向、y方向平均标准差分别为0.1653像素、0.1860像素。基于索引圆环灰度差值最小的光斑中心逼近法在两个对比实验中均得到了最低的平均误差或标准差,具有更高的精度和更好的稳定性,提高了激光告警系统参考位置的定位精度,为光斑中心定位算法提供了新的解决思路。

     

    Abstract: To improve the positioning accuracy of the reference position in the grating diffraction laser warning system, a spot center approximation method based on the minimum grayscale difference of indexed ring is proposed. According to the given initial coordinates of the spot center, a 3x3 center coordinate matrix is generated. Lagrange interpolation is used to calculate the grayscale values of points on the indexed ring centered at the center coordinate matrix. The center coordinate with the minimum grayscale difference on the indexed ring, which represents the true coordinates of the spot center, is approximated using a binary search method. In the accuracy comparison experiment, the proposed algorithm in this paper showed average errors of 0.0013 pixels in the x-direction and 0.0011 pixels in the y-direction across three sets of ideal spot model samples. In the stability comparison experiment, the proposed algorithm in this paper exhibited average standard deviations of 0.1653 pixels in the x-direction and 0.1860 pixels in the y-direction across three sets of actual spot image samples. The spot center approximation method based on the minimum grayscale difference of indexed ring obtained the lowest average error or standard deviation in both comparative experiments. It demonstrates higher accuracy and better stability, thus improving the positioning accuracy of the reference position in the laser warning system. This method provides a new solution for spot center positioning algorithms.

     

/

返回文章
返回