阅读说明：本技术 一种基于平面配准的球面子孔径拼接方法 (Spherical subaperture splicing method based on plane registration ) 是由 朱日宏 秦敏昱 李建欣 马骏 宗毅 段明亮 魏聪 于 2019-10-21 设计创作，主要内容包括：本发明公开了一种基于平面配准的球面子孔径拼接方法,步骤如下：在球面上沿着纬度,每隔一定经度,呈环带规划子孔径拼接路径；将干涉测量得到的子孔径数据,运用Zernike多项式去除低频信息拟合到平面；运用Fourier Mellin图像配准,得到同一环带相邻子孔径在平面内的旋转量和平移量；将同一环带内子孔径依次拼接,在平面内排布,产生二维环带数据；平面环带建立极坐标系,按照极坐标的极径、极角展开,得到矩阵数据；对展开后的矩阵数据进行配准,所有环带和封顶子孔径都在同一平面拼接,得到全口径平面数据；将全口径平面数据转换到球坐标系,形成完整三维球面数据。本发明完成了球面干涉测量中的球面的全口径显示,为球面元件的高精度全表面检测提供参考依据。(The invention discloses a spherical subaperture splicing method based on plane registration, which comprises the following steps: planning sub-aperture splicing paths in an annular band at intervals of certain longitude on a spherical surface along the latitude; removing low-frequency information from the sub-aperture data obtained by interferometry by using a Zernike polynomial and fitting the low-frequency information to a plane; the Fourier Mellin image registration is applied to obtain the rotation amount and the translation amount of the adjacent sub apertures of the same annular zone in the plane; sequentially splicing the subapertures in the same girdle band, and arranging the subapertures in a plane to generate two-dimensional girdle band data; establishing a polar coordinate system by the plane annular belt, and unfolding according to the polar diameter and the polar angle of the polar coordinate to obtain matrix data; registering the expanded matrix data, and splicing all the annular belts and the capping subapertures on the same plane to obtain full-aperture plane data; and converting the full-aperture plane data into a spherical coordinate system to form complete three-dimensional spherical data. The invention completes the full aperture display of the spherical surface in the spherical surface interference measurement and provides a reference basis for the high-precision full surface detection of the spherical element.)

1. A spherical sub-aperture splicing method based on plane registration is characterized by comprising the following steps:

step 1, planning sub-aperture splicing paths in an annular band at intervals of certain longitude along latitude on a spherical surface;

step 2, performing interference measurement on each ring zone according to the sub-aperture splicing path to respectively obtain all sub-aperture data of each ring zone, and removing low-frequency information by using a Zernike polynomial to fit the low-frequency information to a plane;

step 3, conducting Fourier Mellin image registration on each sub-aperture data on the plane to obtain the rotation amount and the translation amount of adjacent sub-apertures of the same annular zone in the plane;

step 4, sequentially splicing the subapertures in the same girdle band, and arranging the subapertures in a plane to generate two-dimensional girdle band data;

step 5, establishing a polar coordinate system according to the two-dimensional annular data, and unfolding according to the polar diameter and the polar angle of the polar coordinate to obtain matrix data;

step 6, registering the matrix data, and splicing all the annular belts and the capping subapertures on the same plane to obtain full-aperture plane data;

and 7, converting the full-aperture plane data into a spherical coordinate system to form complete three-dimensional spherical data.

2. The spherical sub-aperture stitching method based on plane registration as claimed in claim 1, wherein in step 1, the sub-aperture stitching path is planned in a ring zone along the latitude on the spherical surface at regular intervals of longitude, that is, the scanning track of the sub-aperture during measurement is along the latitude line, the sub-aperture acquisition of the next latitude line is continued after the single latitude line is acquired, the sub-apertures in each latitude line ring zone have overlapping parts, the overlapping parts exist among the latitude line rings, and all the sub-apertures can cover the spherical surface with full aperture.

3. The spherical sub-aperture stitching method based on plane registration according to claim 1, wherein in step 2, according to the sub-aperture stitching path, each annular zone is subjected to interferometry, all sub-aperture data of each annular zone are obtained respectively, and Zernike polynomials are used to remove low-frequency information and fit the low-frequency information to a plane, specifically as follows: and (3) according to the sub-aperture splicing path, carrying out interference measurement on each ring band to respectively obtain all sub-aperture data, namely stripe data, of each ring band, carrying out a phase-shifting algorithm on the stripe data to obtain a phase, after the phase is unfolded, removing a low-order wave surface of the sub-aperture by using a Zernike polynomial fitting method, only retaining high-frequency information, and obtaining a plane containing defect information.

4. The spherical sub-aperture stitching method based on plane registration as claimed in claim 1, wherein in step 3, the Fourier Mellin image registration is used to obtain the relative rotation angle Δ θ, the relative displacement Δ x and Δ y of the adjacent sub-apertures in the plane, and the obtained Δ θ, Δ x and Δ y are subjected to error analysis according to the mechanical precision, and data with larger errors are removed and corrected according to the average value.

5. The spherical sub-aperture stitching method based on plane registration according to claim 1, wherein in step 4, the sub-apertures of the same annular zone are stitched, and based on one sub-aperture, the adjacent subsequent sub-apertures are sequentially stitched according to the rotation displacement; all the sub apertures of the annular belt are arranged in a circular arc on a plane; and the last sub-aperture is registered and spliced with the first sub-aperture, so that the two ends of the girdle are the same aperture, and the closed loop is completed on the plane to obtain two-dimensional girdle data.

6. The spherical sub-aperture stitching method based on plane registration as claimed in claim 1, wherein in step 5, according to the two-dimensional zone data, the central coordinates of each sub-aperture in the plane zone are taken, the circle centers of all coordinates are fitted, a polar coordinate system is established with the fitted circle center as a pole and the pole to the first sub-aperture center as a polar axis, and a data point (ρ) is obtained_Pole(s)，θ_Pole(s)) In ρ_Pole(s)Is ordinate, theta_Pole(s)As abscissa, ρ_Pole(s)、θ_Pole(s)Is uniformly distributedAnd rearranging all the data points into the plane rectangle to obtain matrix data.

7. The spherical sub-aperture stitching method based on planar registration as claimed in claim 1, wherein step 6, the planar circular-band diagram is developed according to

8. The spherical sub-aperture stitching method based on plane registration as claimed in claim 1, wherein in step 7, for two-dimensional annular data, the data is divided into (p)_Pole(s)，θ_Pole(s)) As a matrix of coordinates, with the horizontal and vertical coordinates being changed to those in a spherical coordinate system

setting arc line l composed of central points of various sub-apertures of polar coordinate system_kThen the arc is at the angle theta of the spherical coordinate system_{K ball}

Corresponding relationship between other points

Where rho_kThe radius r is equal to l_k/tan(θ_max)，θ_maxThe final sub-aperture central polar angle of the closed loop;

for full-aperture plane data, the horizontal and vertical coordinates of rectangular data are changed into horizontal and vertical coordinates according to corresponding proportional relation

adding the spherical radius R to the data after coordinate conversion to obtain surface type fluctuation R, establishing a spherical coordinate system, substituting the spherical coordinate system into each point

Technical Field

The invention belongs to the field of optical detection, and particularly relates to a spherical sub-aperture splicing method based on plane registration.

Background

The sub-aperture splicing technology is an effective means for detecting the large-aperture optical element with low cost and high resolution. When the size of the measured optical element exceeds the aperture of the interferometer or the density of interference fringes generated by the detection surface is greater than the CCD spatial resolution, the small-aperture interferometer is used for detecting only a part of area (sub-aperture) of the whole optical element each time, and after the full aperture measurement is finished, the full aperture surface information can be obtained by using a proper algorithm for splicing.

In the sub-aperture splicing interferometry method, mechanical positioning errors can affect the alignment of overlapping regions, resulting in errors of relative adjustment coefficients between sub-apertures, thereby affecting the splicing accuracy. At present, the precise positioning method for sub-aperture splicing is mainly realized by improving the precision of a mechanical motion platform, but the method can greatly increase the detection cost and is difficult to achieve the positioning precision of a sub-pixel level. Other methods of assisted positioning complicate the measurement and introduce errors in the camera and the marker itself. The method has the advantages that the mechanical error is compensated and optimized by taking the motion value of the mechanical platform as a reference and then using an algorithm, the method is high in precision, auxiliary measures are not needed, and the method is widely applied.

Lupropylhui adopts a phase-shift diffraction interference target pill full-surface topography detection method based on polarization control in the text of phase-shift diffraction interference target pill full-surface topography detection technology research, provides a sub-aperture data splicing method for basic mapping image matching, maps height information into gray information to generate a mapping image, and then applies an image processing algorithm for matching, thereby greatly reducing the operation amount, improving the processing efficiency of sub-aperture data splicing and shortening the splicing time; but the height information needs to be better processed, so that an image processing algorithm is convenient to match, and the sub-aperture splicing of the spherical surface is perfected.

Disclosure of Invention

The invention aims to provide a spherical sub-aperture splicing method based on plane registration, which completes the full aperture display of a spherical surface in spherical interference measurement and provides a reference basis for high-precision full-surface detection of a spherical element.

The technical solution for realizing the purpose of the invention is as follows: a spherical sub-aperture splicing method based on plane registration comprises the following steps:

step 1, planning sub-aperture splicing paths in an annular band at intervals of certain longitude along latitude on a spherical surface;

step 2, performing interference measurement on each ring zone according to the sub-aperture splicing path to respectively obtain all sub-aperture data of each ring zone, and removing low-frequency information by using a Zernike polynomial to fit the low-frequency information to a plane;

step 3, conducting Fourier Mellin image registration on each sub-aperture data on the plane to obtain the rotation amount and the translation amount of adjacent sub-apertures of the same annular zone in the plane;

step 4, sequentially splicing the subapertures in the same girdle band, and arranging the subapertures in a plane to generate two-dimensional girdle band data;

step 5, establishing a polar coordinate system according to the two-dimensional annular data, and unfolding according to the polar diameter and the polar angle of the polar coordinate to obtain matrix data;

step 6, registering the matrix data, and splicing all the annular belts and the capping subapertures on the same plane to obtain full-aperture plane data;

and 7, converting the full-aperture plane data into a spherical coordinate system to form complete three-dimensional spherical data.

Compared with the prior art, the invention has the remarkable advantages that: (1) the method adopts an image registration algorithm, is applied to the splicing of the sub-apertures, obtains the relative position of each sub-aperture through calculation, and can avoid errors caused by the use of a camera and a mark point compared with a method for reducing the precision of measurement on a mechanical platform by adopting auxiliary measures in the prior art.

(2) By utilizing plane registration splicing, the spherical data is fitted to the plane, so that the influence of surface type errors on registration is reduced. And a complete coordinate conversion system is established, a splicing flow is specified, and plane data is restored to a spherical surface through data processing and conversion on the plane. Moreover, the splicing effect of each step can be visually seen, and the correction can be carried out in real time.

(3) The method can separate the sub-aperture splicing from the dependence on the precision of a mechanical device, can visually detect the effect of each step, and has great market popularization and application values.

Drawings

Fig. 1 is a flowchart of the spherical sub-aperture stitching method based on planar registration.

Fig. 2 is a spherical latitude and longitude scanning track diagram.

Fig. 3 is a plan view after single subaperture fitting.

Fig. 4 is a diagram of the effect of splicing adjacent sub-apertures.

FIG. 5 is a diagram showing the effect of splicing the sub-apertures of the annular band.

FIG. 6 is a diagram of the effect of closed loop on the splicing plane of the sub-apertures of the annular band.

FIG. 7 is a rectangular diagram rearranged by polar expansion.

FIG. 8 is a spherical view of the flat zone after reduction.

Fig. 9 is a diagram showing the effect of inter-zone registration.

Fig. 10 is a diagram of the effect of capping subaperture planar registration.

FIG. 11 is a graph of the spherical effect of the capped sub-aperture and the adjacent annulus.

Fig. 12 is an upper hemisphere effect diagram of the completed planar tiling.

Fig. 13 is a lower hemisphere effect diagram of the completed planar tile.

Fig. 14 is a full aperture plane splicing effect diagram.

Fig. 15 is a full aperture spherical effect diagram.

Detailed Description

The present invention is described in further detail below with reference to the attached drawing figures.

With reference to fig. 1, a spherical sub-aperture stitching method based on planar registration includes the following steps:

step 1, planning a sub-aperture splicing path in an annular band at intervals of a certain longitude along a latitude on a spherical surface:

the sub-aperture scanning track during measurement is along the wefts, the sub-aperture collection of the next weft is continued after the single weft is completely collected, the sub-apertures in each weft ring zone have a superposition part, the superposition parts also exist among the weft ring zones, and all the sub-apertures can cover a full-aperture spherical surface. The overlap ratio of adjacent sub-apertures is about 25%, too large an overlap ratio increases the measurement engineering amount, and too small an overlap ratio may cause too large an error.

Step 2, performing interference measurement on each ring zone according to the sub-aperture splicing path to respectively obtain all sub-aperture data of each ring zone, and removing low-frequency information by using a Zernike polynomial to fit to a plane:

according to the subaperture splicing path, carrying out interference measurement on each annular band to respectively obtain all subaperture data of each annular band, and removing low-frequency information by using a Zernike polynomial to fit to a plane, wherein the details are as follows: and (3) according to the sub-aperture splicing path, carrying out interference measurement on each ring band to respectively obtain all sub-aperture data, namely stripe data, of each ring band, carrying out a phase-shifting algorithm on the stripe data to obtain a phase, after the phase is unfolded, removing a low-order wave surface of the sub-aperture by using a Zernike polynomial fitting method, only retaining high-frequency information, and obtaining a plane containing defect information. The height information is mapped into gray information to generate a mapping image, and then an image processing algorithm is applied to carry out matching, so that the calculation amount is greatly reduced, the processing efficiency of sub-aperture data splicing is improved, and the splicing time is shortened.

Step 3, performing Fourier Mellin image registration on each sub-aperture data on the plane to obtain the rotation amount and the translation amount of the adjacent sub-apertures of the same annular zone in the plane:

and obtaining the relative rotation angle delta theta, the relative displacement delta x and the relative displacement delta y of adjacent sub-apertures in a plane through Fourier Mellin image registration, carrying out error analysis on the obtained delta theta, delta x and delta y according to mechanical precision, removing data with larger errors and correcting according to an average value. The Fourier Mellin image registration method has small error, but needs to pay attention to the fact that a proper threshold value is adopted during operation, and efficiency can be improved.

And 4, sequentially splicing the subapertures in the same girdle band, and arranging in a plane to generate two-dimensional girdle band data:

splicing the sub-apertures of the same annular belt, and sequentially splicing the adjacent subsequent sub-apertures according to the rotary displacement by taking one sub-aperture as a reference; all the sub apertures of the annular belt are arranged in a circular arc on a plane; and the last sub-aperture is registered and spliced with the first sub-aperture, so that the two ends of the girdle are the same aperture, the closed loop is completed on the plane, the two-dimensional girdle data is obtained, and each defect pattern can be visually seen from the two-dimensional girdle data to see the splicing effect.

Step 5, establishing a polar coordinate system according to the two-dimensional annular data, and unfolding according to the polar diameter and the polar angle of the polar coordinate to obtain matrix data:

according to the two-dimensional annular data, the central coordinates of each sub-aperture in the planar annular band are taken, the circle centers of all coordinates are fitted, a polar coordinate system is established by taking the fitted circle center as a pole and the pole to the center of the first sub-aperture as a polar axis, and a data point (rho) is obtained_Pole(s)，θ_Pole(s)) In ρ_Pole(s)Is ordinate, theta_Pole(s)As abscissa, ρ_Pole(s)、θ_Pole(s)And uniformly distributing, and rearranging all data points into a plane rectangle to obtain matrix data. The matrix data is a deformation of the two-dimensional zone data, and the defect patterns on their surfaces all follow the same transformation.

Step 6, registering the matrix data, and splicing all the annular bands and the capping subapertures on the same plane to obtain full-aperture plane data:

according to the plan view after deployment, i.e.Theta coordinate matrix, and phase correlation technique for inter-zone registration to obtain inter-zone displacement

Delta theta is the deviation between the circular bands in the spherical coordinate system; by joining adjacent zones in a plane

And (3) splicing theta coordinate data, splicing adjacent annular belts on the same plane in sequence according to the method, and unfolding, registering and splicing the capped sub-apertures according to the same steps to finally complete the plane full-aperture data. The complete defect pattern can be directly seen from the plane full aperture data, and the defect pattern can be properly corrected.

And 7, converting the full-aperture plane data into a spherical coordinate system to form complete three-dimensional spherical data, which is as follows:

for two-dimensional zone data, will be represented by (ρ)_Pole(s)，θ_Pole(s)) As a matrix of coordinates, with the horizontal and vertical coordinates changed to those in a spherical coordinate system

θ_{Ball with ball-shaped section}) The corresponding relationship is as follows:

setting arc line l composed of central points of various sub-apertures of polar coordinate system_kThen the arc is at the angle theta of the spherical coordinate system_{K ball}

Corresponding relationship between other points

Where rho_kIs a radius of a pole corresponding to a central arc line in a polar coordinate systemDiameter r ═ l_k/tan(θ_max)，θ_maxThe final sub-aperture central polar angle of the closed loop;

for full-aperture plane data, changing the horizontal and vertical coordinates of rectangular data into (A) θ_{Ball with ball-shaped section}) Coordinates of

Theta is uniformly distributed in the range of 0-2 pi and becomes a spherical coordinate;

the data after coordinate conversion is added with the spherical radius R to obtain the surface undulation R, a spherical coordinate system is established, and the spherical coordinates (R,θ_{ball with ball-shaped section}) And forming three-dimensional spherical full-aperture data. By rotating the obtained three-dimensional spherical surface, a defective pattern such as a peak can be observed and analyzed.