高分辨LCOPA相移分布自动检测方法研究
发布时间：2018-12-31 11:48
【摘要】：随着激光技术的快速发展,液晶光学相控阵(LCOPA)已经成为波控技术中的代表之一,作为一种实时可编程的电控波束扫描技术,具有其他技术无可匹敌的优势:功耗低、重量轻、控制灵活等。但其在最大偏转角度、效率以及有效孔径等方面均有明显不足。目前的相位差检测只是针对均匀相控器件,而有效检测液晶相控阵器件的二维相位差分布情况,可以为其进一步发展或应用提供依据。本论文以液晶相控阵器件的二维相位差高分辨检测为目的,利用常规四分之一波片法的检测优势,研究光路中数据采集技术以及正弦曲线形式的数据处理算法,提出了傅里叶四分之一波片法。该检测法对1/4波片、液晶相控阵器件进行了实验分析,实现了二维相控器件的二维高分辨检测。创新性地将图像处理算法运用于二维相位差的匹配以及融合过程中,解决了液晶相控阵器件的大视场检测问题。主要研究内容如下:1)基于常规四分之一波片法,通过添加显微物镜、CCD、步进电机等实现检测自动化,并研究了DFT算法计算有效频率分量的最小值,提出了傅里叶四分之一波片法。并对检测装置进行三维立体设计,提高检测可靠性。2)以最小二乘法为基本原理的拟合算法与离散傅里叶算法在理论、仿真、检测实验上进行了对比分析,离散傅里叶变换算法(DFT)具有较高的效率与精度。运用傅里叶四分之一波片法分别对1/4波片、液晶相控阵器件进行了检测实验,1/4波片法的平均检测误差小于0.3°,检测液晶相控阵的二维相位差可得到3μm电极、间隔子以及相位凹陷的具体影响。从而验证了傅里叶四分之一波片法在二维相位差高分辨检测上的有效性。3)运用图像处理算法中的Hough变换算法、Radon变换算法、线性回归算法,对二维相位差的倾斜角度问题进行了研究。理论上分析了临近插值算法、二维插值算法、双三次插值算法实现相位差分布的无失真纠斜问题。针对二维相位差数据的拼接问题,分别分析了基于区域的匹配算法与基于Harris特征的匹配算法,解决匹配位置的计算问题,然后对比中值平均法等,采用了加权平均法对重叠区域的数据融合处理,实现了拼接中的无缝合成。
[Abstract]:With the rapid development of laser technology, liquid crystal optical phased array (LCOPA) has become one of the representative of wave control technology. As a real-time programmable electronic control beam scanning technology, it has the unparalleled advantages of other technologies: low power consumption. Light weight, flexible control, etc. However, its maximum deflection angle, efficiency and effective aperture are obviously insufficient. The current phase difference detection is only for homogeneous phase control devices, but the effective detection of the two-dimensional phase difference distribution of liquid crystal phased array devices can provide a basis for its further development or application. The purpose of this paper is to detect the two-dimensional phase difference high resolution of liquid crystal phased array devices, and to study the data acquisition technology and sinusoidal data processing algorithm based on the conventional 1/4 wave plate method. Fourier 1/4 wave plate method is proposed. The method is used to analyze the 1 / 4 wave plate and liquid crystal phased array devices, and the two-dimensional high resolution detection of the two dimensional phase control devices is realized. The image processing algorithm is innovatively applied to the matching and fusion of two-dimensional phase difference, which solves the problem of large field of view detection of liquid crystal phased array devices. The main research contents are as follows: 1) based on the conventional 1/4 wave plate method, the detection automation is realized by adding micro objective lens and CCD, step motor, and the minimum value of effective frequency component is calculated by DFT algorithm. Fourier 1/4 wave plate method is proposed. In order to improve the reliability of the detection system, 3D design is carried out. 2) the fitting algorithm based on the least square method and the discrete Fourier algorithm are compared and analyzed in theory, simulation and test experiment. The discrete Fourier transform (DFT) algorithm has high efficiency and accuracy. The 1 / 4 wave plate and liquid crystal phased array device are tested by Fourier 1/4 wave plate method. The average detection error of 1 / 4 wave plate method is less than 0.3 掳, and the 3 渭 m electrode can be obtained by detecting the two-dimensional phase difference of liquid crystal phased array. The specific effects of spacers and phase indentation. Thus, the validity of Fourier 1/4 wave plate method in high resolution detection of two-dimensional phase difference is verified. 3) the Hough transform algorithm, Radon transform algorithm and linear regression algorithm are used in image processing algorithm. The tilting angle of two-dimensional phase difference is studied. In this paper, we analyze approximate interpolation algorithm, two-dimensional interpolation algorithm and bicubic interpolation algorithm to realize the distortion correction problem of phase difference distribution. In order to solve the problem of two dimensional phase difference data splicing, the region based matching algorithm and the Harris feature based matching algorithm are analyzed respectively to solve the problem of calculating the matching position, and then the median average method is compared, and so on. The method of weighted average is used to deal with the data fusion of overlapped regions, and the seamless composition in stitching is realized.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41;TN249
本文编号：2396530
[Abstract]:With the rapid development of laser technology, liquid crystal optical phased array (LCOPA) has become one of the representative of wave control technology. As a real-time programmable electronic control beam scanning technology, it has the unparalleled advantages of other technologies: low power consumption. Light weight, flexible control, etc. However, its maximum deflection angle, efficiency and effective aperture are obviously insufficient. The current phase difference detection is only for homogeneous phase control devices, but the effective detection of the two-dimensional phase difference distribution of liquid crystal phased array devices can provide a basis for its further development or application. The purpose of this paper is to detect the two-dimensional phase difference high resolution of liquid crystal phased array devices, and to study the data acquisition technology and sinusoidal data processing algorithm based on the conventional 1/4 wave plate method. Fourier 1/4 wave plate method is proposed. The method is used to analyze the 1 / 4 wave plate and liquid crystal phased array devices, and the two-dimensional high resolution detection of the two dimensional phase control devices is realized. The image processing algorithm is innovatively applied to the matching and fusion of two-dimensional phase difference, which solves the problem of large field of view detection of liquid crystal phased array devices. The main research contents are as follows: 1) based on the conventional 1/4 wave plate method, the detection automation is realized by adding micro objective lens and CCD, step motor, and the minimum value of effective frequency component is calculated by DFT algorithm. Fourier 1/4 wave plate method is proposed. In order to improve the reliability of the detection system, 3D design is carried out. 2) the fitting algorithm based on the least square method and the discrete Fourier algorithm are compared and analyzed in theory, simulation and test experiment. The discrete Fourier transform (DFT) algorithm has high efficiency and accuracy. The 1 / 4 wave plate and liquid crystal phased array device are tested by Fourier 1/4 wave plate method. The average detection error of 1 / 4 wave plate method is less than 0.3 掳, and the 3 渭 m electrode can be obtained by detecting the two-dimensional phase difference of liquid crystal phased array. The specific effects of spacers and phase indentation. Thus, the validity of Fourier 1/4 wave plate method in high resolution detection of two-dimensional phase difference is verified. 3) the Hough transform algorithm, Radon transform algorithm and linear regression algorithm are used in image processing algorithm. The tilting angle of two-dimensional phase difference is studied. In this paper, we analyze approximate interpolation algorithm, two-dimensional interpolation algorithm and bicubic interpolation algorithm to realize the distortion correction problem of phase difference distribution. In order to solve the problem of two dimensional phase difference data splicing, the region based matching algorithm and the Harris feature based matching algorithm are analyzed respectively to solve the problem of calculating the matching position, and then the median average method is compared, and so on. The method of weighted average is used to deal with the data fusion of overlapped regions, and the seamless composition in stitching is realized.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41;TN249
【参考文献】
相关期刊论文 前2条
1 蔡胜利;张会清;;基于Harris角点检测的图像旋转测量[J];计算机测量与控制;2011年01期
2 苏美开;宋连科;李艺;李国华;;波片测量中的补偿法[J];应用激光;1990年05期
，本文编号：2396530
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2396530.html
