1. A precision workpiece stage inclination angle detection method based on image characteristic information is characterized by comprising the following steps:

s1: the method comprises the steps of placing a calibration object on a precision workbench in a horizontal state in advance, acquiring images to obtain a standard image, and then placing the calibration object on the precision workbench and acquiring the images to obtain a current sampling image when the precision workbench needs to perform inclination angle detection;

s2: respectively extracting and matching the characteristic points of the standard graph and the sampling graph to obtain a characteristic point pair set;

s3: selecting three pairs of feature point pairs from feature point pair sets of the standard graph and the sampling graph, and recording the coordinates of each feature point in the image asi-0 represents that the characteristic point belongs to the standard graph, and i-1 represents that the characteristic point belongs to the sampling graph;

s4: calculating the inclination angle of the precision workbench according to the three pairs of characteristic point pairs obtained in the step S3, comprising the following steps:

s4.1: three feature points belonging to the standard graph and the sampling graph in the three pairs of feature point pairs are respectively used as vertexes to form a feature triangle delta A⁰B⁰C⁰、ΔA¹B¹C¹Then respectively calculating the side lengths of the characteristic triangles And

s4.2: establishing a space rectangular coordinate system by taking the original point of a plane rectangular coordinate system of a sampling diagram as the original point and the sampling diagram as the horizontal plane, and enabling a feature triangle delta A⁰B⁰C⁰Move to make the feature point A⁰And feature point A¹Coincide and then at a characteristic point A⁰For characteristic triangle delta A under fixed conditions⁰B⁰C⁰Zooming to obtain a feature triangleLet a characteristic triangleThe vertical projection on the horizontal plane is a characteristic triangle delta A¹B¹C¹， Respectively represent the characteristic triangle Δ A⁰B⁰C⁰Moving and zooming the characteristic point A⁰Characteristic point B⁰And a characteristic point C⁰And (3) corresponding points, taking the scaling factor x as an unknown parameter, and constructing the following equation set:

wherein the content of the first and second substances,indicating pointsThe value of the height relative to the horizontal plane,indicating pointsA height value relative to a horizontal plane;

solving the equation set to obtain two sets of height values, which are respectively recorded as

S4.3: two groups of height values are respectively adopted to calculate to obtain two groups of precise workbench inclination angles, and the calculation method is as follows:

noting that the currently used height value isk is 1,2, and a characteristic triangle delta A in the sampling diagram¹B¹C¹The coordinates of the three characteristic points in the established space rectangular coordinate system are respectively Then characteristic triangleThe coordinates of the three characteristic points in the space rectangular coordinate system are respectively

In a space rectangular coordinate system, calculating a normal vector perpendicular to a plane where a standard diagram feature triangle is locatedAccording to the normal vectorCalculating to obtain the inclination angle of the precision workbench in the x-axis direction in a space rectangular coordinate systemAnd inclination in the y-axis direction

2. The method of claim 1, wherein the step of selecting the characteristic point pairs in step S3 comprises the steps of:

s3.1: recording the number of the characteristic point pairs in the characteristic point pair set as N, and arranging the N pairs of characteristic point pairs from large to small according to the matching degree of the two characteristic points;

s3.2: selecting a 1 st pair of characteristic point pairs from the arranged characteristic point pair sequence as a characteristic point pair A;

s3.3: let n be 2;

s3.4: judging the distance d between the n-th pair of characteristic point pairs and the characteristic point pair A_nAWhether the side length is larger than a preset side length threshold value D or not is judged, if not, the step S3.5 is carried out, otherwise, the step S3.6 is carried out;

s3.5: making n equal to n +1, and returning to step S3.4;

s3.6: taking the current characteristic point pair as a characteristic point pair B;

s3.7: let n ═ n + 1;

s3.8: judgment ofThe nth' pair of characteristic point pairs satisfies the condition of the characteristic point pair C, that is, satisfies the following three conditions at the same time: the distance d between the n' th pair of characteristic point pairs and the characteristic point pair A_n′AThe distance D between the n' th pair of characteristic point pairs and the characteristic point pair B is larger than a preset side length threshold D_n′BGreater than a predetermined side length threshold D, D_n′A+d_n′B-d_A,B> P, wherein d_A,BRepresenting the distance between the characteristic point pair A and the characteristic point pair B, wherein P is a preset threshold value, if not, entering a step S3.9, otherwise, entering a step S3.10;

s3.9: making n '═ n' +1, return to step S3.8;

s3.10: and taking the current characteristic point pair as a characteristic point pair C.

Background

For many precision machines, adjusting the workpiece stage to a level position is a critical step prior to subsequent operations. If the workpiece stage of the photoetching machine is not adjusted to be in a horizontal state, the surface of the silicon wafer and the focal plane of the projection objective lens are in a non-parallel state, so that the etching precision of the silicon wafer is seriously influenced, and the quality and the yield of a chip manufacturing product are reduced. Meanwhile, methods for measuring flatness, such as a surface measuring method and a beam plane method, which are common in machinery, often have their own measurement limitations, and the measurement accuracy is not high.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provide a precise workpiece stage inclination angle detection method based on image characteristic information.

In order to achieve the above object, the method for detecting the inclination angle of the precision workpiece stage based on the image characteristic information comprises the following steps:

s1: the method comprises the steps of placing a calibration object on a precision workbench in a horizontal state in advance, acquiring images to obtain a standard image, placing the calibration object on the precision workbench when the precision workbench needs to perform inclination angle detection, and acquiring the images to obtain a current sampling image;

s2: respectively extracting and matching the characteristic points of the standard graph and the sampling graph to obtain a characteristic point pair set;

s3: selecting three pairs of characteristic point pairs from characteristic point pair sets of the standard graph and the sampling graph, and recording the positions of the characteristic pointsThe coordinates in the image are respectivelyi-0 represents that the characteristic point belongs to the standard graph, and i-1 represents that the characteristic point belongs to the sampling graph;

s4: calculating the inclination angle of the precision workbench according to the three pairs of characteristic point pairs obtained in the step S3, comprising the following steps:

s4.1: respectively taking three feature points belonging to the standard graph and the sampling graph in the three pairs of feature point pairs as vertexes to form a feature triangle delta A⁰B⁰C⁰、ΔA¹B¹C¹Then respectively calculating the side lengths of the characteristic triangles And

s4.2: establishing a space rectangular coordinate system by taking the original point of a plane rectangular coordinate system of a sampling diagram as the original point and the sampling diagram as the horizontal plane, and enabling a feature triangle delta A⁰B⁰C⁰Move to make the feature point A⁰And feature point A¹Coincide and then at a characteristic point A⁰For characteristic triangle delta A under fixed conditions⁰B⁰C⁰Zooming to obtain a feature triangleLet a characteristic triangleThe vertical projection on the horizontal plane is a characteristic triangle delta A¹B¹C¹， Respectively represent the characteristic triangle Δ A⁰B⁰C⁰Moving and zooming the characteristic point A⁰Characteristic point B⁰And a characteristic point C⁰And (3) corresponding points, taking the scaling factor x as an unknown parameter, and constructing the following equation set:

wherein the content of the first and second substances,indicating pointsThe value of the height relative to the horizontal plane,indicating pointsA height value relative to a horizontal plane;

solving the equation set to obtain two sets of height values, which are respectively recorded as

S4.3: two groups of height values are respectively adopted to calculate to obtain two groups of precise workbench inclination angles, and the calculation method is as follows:

noting that the currently used height value isk is 1,2, and a characteristic triangle delta A in the sampling diagram¹B¹C¹The coordinates of the three characteristic points in the established space rectangular coordinate system are respectively Then characteristic triangleThe coordinates of the three characteristic points in the space rectangular coordinate system are respectively

In a space rectangular coordinate system, calculating a normal vector perpendicular to a plane where a standard diagram feature triangle is locatedAccording to the normal vectorCalculating to obtain the inclination angle of the precision workbench in the x-axis direction in the space rectangular coordinate systemAnd inclination in the y-axis direction

The invention relates to a precision workpiece table inclination angle detection method based on image characteristic information, which comprises the steps of placing a calibration object on a precision working table in a horizontal state in advance, acquiring an image to obtain a standard image, placing the calibration object on the precision working table, acquiring the image to obtain a current sampling image when the precision working table needs to be subjected to inclination angle detection, respectively extracting and matching characteristic points of the standard image and a sampling image to obtain a characteristic point pair set, selecting three pairs of characteristic point pairs from the characteristic point pair set of the standard image and the sampling image, and calculating according to a designed inclination angle calculation method to obtain the inclination angle of the precision working table.

The invention completes the non-contact type precision workbench inclination angle detection by utilizing the acquired image information, can effectively save the structural hardware cost, provides a new inclination angle calculation method and improves the precision workbench inclination angle detection accuracy.

Drawings

FIG. 1 is a flow chart of an embodiment of the method for detecting the tilt angle of a precision workpiece stage based on image feature information according to the present invention;

fig. 2 is a flowchart of a feature point pair selection method in the present embodiment;

FIG. 3 is a positional relationship diagram of a sampling diagram and a standard diagram in a three-dimensional space according to the present invention;

FIG. 4 is a flow chart of the calculation of the tilt angle of the precision stage in the present invention;

FIG. 5 is a schematic diagram of the relative positions of a sampling map and a standard map;

FIG. 6 is a schematic diagram of scaling of feature triangles in a standard graph;

FIG. 7 is a schematic diagram of the normal vectors of the feature triangles shown in FIG. 6;

fig. 8 is an exemplary diagram of feature triangles in the sample graph and the standard graph in the present embodiment.

Detailed Description

The following description of the embodiments of the present invention is provided in order to better understand the present invention for those skilled in the art with reference to the accompanying drawings. It is to be expressly noted that in the following description, a detailed description of known functions and designs will be omitted when it may obscure the subject matter of the present invention.

Examples

FIG. 1 is a flowchart of an embodiment of a method for detecting an inclination angle of a precision workpiece stage based on image feature information according to the present invention. As shown in fig. 1, the method for detecting the tilt angle of the precision workpiece stage based on the image feature information of the present invention specifically comprises the following steps:

s101: image acquisition:

the method comprises the steps of placing a calibration object on a precision workbench in a horizontal state in advance, acquiring images to obtain a standard image, and then placing the calibration object on the precision workbench and acquiring the images to obtain a current sampling image when the precision workbench needs to perform inclination angle detection.

S102: matching the characteristic points:

and respectively extracting and matching the characteristic points of the standard graph and the sampling graph to obtain a characteristic point pair set.

In this embodiment, before feature point extraction is performed on the standard graph and the sampling graph, the two images may be preprocessed, for example, by filtering and sharpening, then feature point extraction is performed on the preprocessed two images by using orb (organized Fast and Rotated brief) operators, and matching of feature points is completed by using a FLANN (Fast approximation Nearest Neighbor Search Library) algorithm.

S103: selecting characteristic point pairs:

selecting three pairs of characteristic point pairs from characteristic point pair sets of the standard graph and the sampling graph, and recording the coordinates of each characteristic point in the image asi-0 indicates that the feature point belongs to the standard graph, and i-1 indicates that the feature point belongs to the sampling graph.

In order to make the final detected precision workbench inclination more accurate, the feature point pair for inclination calculation preferably satisfies the following two points: firstly, designing a screening calculation method according to the theorem that the sum of two sides of a triangle is greater than the third side, and avoiding the situation that the obtuse angle is close to a straight angle in order to improve the calculation precision of the inclination angle; secondly, the distance between the feature point pairs cannot be too close, because the calculation of the vertices of the feature triangle in this case affects the accuracy of the dip angle value calculation. In summary, the characteristic point pair selection method of the embodiment is designed. Fig. 2 is a flowchart of a feature point pair selecting method in this embodiment. As shown in fig. 2, the specific method for selecting the characteristic point pairs in this embodiment is as follows:

s201: and (3) arranging characteristic point pairs:

and counting the number of the characteristic point pairs in the characteristic point pair set as N, and arranging the N pairs of characteristic point pairs according to the matching degree of the two characteristic points from large to small.

S202: determining a characteristic point pair A:

and selecting the 1 st pair of characteristic point pairs from the arranged characteristic point pair sequence as the characteristic point pair A.

S203: let n be 2.

S204: judging the distance d between the n-th pair of characteristic point pairs and the characteristic point pair A_nAAnd if not, the step S205 is carried out, otherwise, the step S206 is carried out.

S205: let n be n +1, return to step S204.

S206: determining a characteristic point pair B:

and taking the current characteristic point pair as a characteristic point pair B.

S207: let n ═ n + 1;

s208: judging whether the n' th pair of characteristic point pairs meets the condition of the characteristic point pair C, namely whether the following three conditions are met simultaneously: the distance d between the n' th pair of characteristic point pairs and the characteristic point pair A_n′AThe distance D between the n' th pair of characteristic point pairs and the characteristic point pair B is larger than a preset side length threshold D_n′BGreater than a predetermined side length threshold D, D_n′A+d_n′B-d_A,B> P, wherein d_A,BAnd representing the distance between the characteristic point pair A and the characteristic point pair B, wherein P is a preset threshold, if not, the step S209 is carried out, otherwise, the step S210 is carried out.

S209: let n' +1, return to step S208.

S210: and determining a characteristic point pair C:

and taking the current characteristic point pair as a characteristic point pair C.

S104: calculating the inclination angle:

and calculating the inclination angle of the precision workbench according to the three pairs of characteristic points obtained in the step S103.

Knowing that three points which are not collinear define a plane, a method of locating the plane is obtained, i.e.A single plane can be located at least by the coordinates of a spatial rectangular coordinate system of three points which are not in a line. In the invention, the sampling graph is taken as a horizontal plane, and the standard graph is taken as an inclined plane. FIG. 3 is a positional relationship diagram of a sampling chart and a standard chart in a three-dimensional space in the present invention. As shown in FIG. 3, the coordinate values of the six points in the rectangular space coordinate system are combined with the knowledge about the spatial solid geometry to determine the inclined plane A⁰B⁰C⁰The normal quantity of the precision workbench, namely the inclined plane is characterized, so that the inclination angle of the precision workbench is obtained.

FIG. 4 is a flow chart of the calculation of the tilt angle of the precision stage in the present invention. As shown in fig. 4, the specific steps of the calculation of the inclination angle of the precision worktable in the present invention include:

s401: determining a characteristic triangle:

respectively forming a characteristic triangle delta A by three characteristic points belonging to the standard graph and the sampling graph in the three pairs of characteristic points⁰B⁰C⁰、ΔA¹B¹C¹Then respectively calculating the side lengths of the characteristic trianglesAnd

s402: calculating the height of the vertex of the characteristic triangle:

and establishing a space rectangular coordinate system by taking the original point of the plane rectangular coordinate system of the sampling graph as the original point and the sampling graph as the horizontal plane, and then calculating the height of the vertex of the characteristic triangle in the standard graph relative to the horizontal plane.

As shown in fig. 3, in step S103, the coordinates of each feature point in the two-dimensional rectangular coordinate system of each image are obtained, and the length of three sides of two feature triangles in each image is calculated. The characteristic triangle Delta A in the sampling picture¹B¹C¹And feature triangles in standard graphShape Delta A⁰B⁰C⁰And translating to the same space rectangular coordinate system, and enabling the two characteristic points of the characteristic point pair A in the two characteristic triangles to be intersected. Fig. 5 is a schematic diagram of the relative positions of the sampling map and the standard map. As shown in fig. 5, the relative positions of the sample chart and the standard chart are four cases as shown in fig. 5(a), 5(b), 5(c), and 5 (d). To facilitate observation, a dotted line is added to the sample and standard images to characterize the plane. According to the principle of image acquisition, Delta A¹B¹C¹Is Δ A⁰B⁰C⁰Vertical projection in the horizontal plane.Are respectively a characteristic point B⁰And characteristic point C⁰Relative to a characteristic point B on a horizontal plane¹And characteristic point C¹Wherein the arrow direction indicates the height direction of the inclined surface at the point with respect to the horizontal plane, the arrow direction being upward indicates that the point is above the horizontal plane, and the arrow direction being downward indicates that the point is below the horizontal plane.

As can be seen from FIG. 5, at characteristic point A⁰、A¹Under the condition of coincidence, the spatial positional relationship between the sample map and the standard map can be generally divided into two cases. The first case is that in a rectangular spatial coordinate system, the characteristic point A¹(A⁰) The height position in the vertical direction is at the feature point B¹And characteristic point C¹On the same side (see fig. 5(a) and 5(b)), and the second case is the feature point a¹(A⁰) The height position in the vertical direction is at the feature point B¹Characteristic point C¹Between the height positions (see fig. 5(c), 5 (d)). The situation is again classified as feature point A in FIG. 5(a)¹(A⁰) While being higher than the characteristic point B¹And characteristic point C¹And a characteristic point A of FIG. 5(b)¹(A⁰) While being lower than the characteristic point B¹And characteristic point C¹Two cases of points. The situation is divided into two as the characteristic point C of FIG. 5(C)⁰Is the highest point, and the characteristic point B of FIG. 5(d)⁰The two cases of the highest point.

In the actual image capturing process, when the workpiece stage is in the inclined state and the horizontal state, the distance between the calibration object placed on the workpiece stage and the photoelectric sensor changes, so that the pixel distance between two identical feature points on the calibration object on the image also changes, that is, the feature point B on the standard graph changes⁰Or characteristic point C⁰The vertical projection of (B) may not fall on the feature point B of the sample map¹Or characteristic point C¹This will seriously affect the results of the tilt calculation. However, although the pixel distance between two points changes, the proportion of the length of three sides of the triangle determined by the characteristic point is fixed, and the triangle of the standard graph can be reduced or enlarged in the same plane by a certain proportion to ensure that the vertical projection of the three points just coincides with the triangle of the sampling graph.

FIG. 6 is a schematic diagram of scaling of feature triangles in a standard graph. As shown in FIG. 6, for the case as in FIG. 5(a), triangle Δ A⁰B⁰C⁰Is the initial position of the feature triangle on the standard graph by dividing one side B⁰C⁰The translation is carried out in the plane of the standard graph, and the position where the vertical projection of the characteristic point in the standard graph is coincident with the characteristic point in the sampling graph can be determined and is unique.

Let the characteristic triangle delta A⁰B⁰C⁰Make translation to make characteristic point A⁰And feature point A¹Coincide and then at a characteristic point A⁰For characteristic triangle delta A under fixed conditions⁰B⁰C⁰Zooming to obtain a feature triangleWhen the scaling factor is x, the following expression is given:

wherein the content of the first and second substances,respectively representing scaled feature trianglesMiddle triangle sideThe length of the side of the frame,respectively represent the characteristic triangle Δ A⁰B⁰C⁰The feature point A after zooming⁰Characteristic point B⁰And a characteristic point C⁰The corresponding point.

Due to the characteristic triangle delta A¹B¹C¹Is a scaled feature triangleFor vertical projection in the horizontal plane, there are

In summary, the feature triangle Δ A is used⁰B⁰C⁰Move to make the feature point A⁰And feature point A¹Coincide and then at a characteristic point A⁰For characteristic triangle delta A under fixed conditions⁰B⁰C⁰Zooming to obtain a feature triangleLet a characteristic triangleThe vertical projection on the horizontal plane is a characteristic triangle delta A¹B¹C¹， Respectively represent the characteristic triangle Δ A⁰B⁰C⁰Moving and zooming the characteristic point A⁰Characteristic point B⁰And a characteristic point C⁰And (3) corresponding points, taking the scaling factor x as an unknown parameter, and constructing the following equation set:

wherein the content of the first and second substances,indicating pointsThe value of the height relative to the horizontal plane,indicating pointsHeight values relative to the horizontal plane.

Solving the equation set to obtain two sets of height values, which are respectively recorded asThese two sets of height values correspond to the two results of (a) and (b) in fig. 5 (or (c) and (d) in fig. 5), respectively. Since the vertical projection imaging results of the calibration objects in the two tilted states are the same, the solution of the equation set cannot be directly selected through the image information. The selection of the result can be performed by a trial and error method.

S403: calculating the inclination angle:

two groups of height values are respectively adopted to calculate to obtain two groups of precise workbench inclination angles, and the calculation method is as follows:

noting that the currently used height value isk is 1,2, and a characteristic triangle delta A in the sampling diagram¹B¹C¹The coordinates of the three characteristic points in the established space rectangular coordinate system are respectively Then characteristic triangleThe coordinates of the three characteristic points in the space rectangular coordinate system are respectively

In a space rectangular coordinate system, calculating a triangle perpendicular to the characteristicNormal vector of the planeFig. 7 is a schematic diagram of normal vectors of the feature triangles shown in fig. 6. In actual calculation, a single-bit normal vectorTwo sets of solutions are generated, corresponding to the normal vectors of the inclined surface to be measured facing upwards and downwards respectively, and the two have no difference in the subsequent inclination angle calculation results, and only the solution is takenFor subsequent use.

I.e. according to the normal vectorCalculating to obtain the inclination angle of the precision workbench in the x-axis direction in a space rectangular coordinate systemAnd inclination in the y-axis direction

In order to better illustrate the technical scheme of the invention, the invention is experimentally verified by adopting specific examples. The inclination angle of the precision workpiece table in the x-axis direction is preset to be 5 degrees, and the inclination angle in the y-axis direction is-5 degrees.

Fig. 8 is an exemplary diagram of feature triangles in the sample graph and the standard graph in the present embodiment. As shown in fig. 8, the feature point coordinates of the three vertices of the feature triangle in the sampling diagram are:

A¹(146.68,494.979),B¹(82.1146,364.66),C¹(232.707,427.019)

the coordinates of three feature points matched with the standard graph are as follows:

A⁰(594.808,498.062),B⁰(530.012,373.248),C⁰(677.221,429.982)

obtaining a characteristic triangle delta A in the sampling graph from the characteristic point coordinates¹B¹C¹The side lengths of are respectively:

B¹C¹＝162.993,A¹C¹＝109.632,A¹B¹＝145.436

feature triangles in standard drawingsΔA⁰B⁰C⁰The side length of (A) is:

B⁰C⁰＝157.763,A⁰C⁰＝106.896,A⁰B⁰＝140.631

the system of equations in step S402 is solved to obtain two sets of solutions. Selection in this example This set of solutions. Characteristic triangle delta A in sampling chart¹B¹C¹The coordinates of the three characteristic points in the spatial rectangular coordinate system are respectively as follows:

A¹(146.68,494.979,0),B¹(82.1146,364.66,0),C¹(232.707,427.019,0)

then characteristic triangleThe coordinates of the three characteristic points in the space rectangular coordinate system are respectively

Calculating to obtain a feature triangleNormal vector of (1)Has a value ofFrom the normal vectorThe inclination angle theta of the inclined plane in the x-axis direction and the y-axis direction respectively can be calculated_x＝5.71745°，θ_y＝-5.18507°。

In order to illustrate the effectiveness of the invention, two additional sets of feature point pairs are selected for re-tilt calculation.

Table 1 is a table of calculated tilt statistics for three groups of feature point pairs in this embodiment.

TABLE 1

Therefore, the error of the calculated inclination angle and the actual inclination angle is within the acceptable range of engineering application, and the effectiveness of the invention is illustrated.

In addition, the invention also provides a screening method of the characteristic point pairs, so the technical effects of randomly selecting the characteristic point pairs and screening and determining the characteristic point pairs are compared. And selecting three groups of characteristic point pairs for dip angle calculation in each mode. Table 2 is a table comparing the results of the dip calculations for randomly selected pairs of feature points and for pairs of feature points determined by screening in the present invention.

TABLE 2

As shown in table 2, when the feature point pair is randomly selected, the obtained final result of the tilt angle is greatly different from the actual tilt angle. Specifically, analyzing the data in table 5-2, it can be seen that the calculated tilt angle results of the first and third sets of feature triangle pairs of randomly selected feature point pairs are closer to the actual tilt angle value, but the calculated results of the second set are severely deviated. The coordinate information of the feature point pair corresponding to the second group of data is found in the backtracking experiment process, and the position of two feature points in the image is found to be too tight, namely the length of one side of the obtained feature triangle is too small, so that the inclination angle calculation result is greatly influenced. And the characteristic points are determined by screening to be more stable to the obtained inclination angle error.

Although illustrative embodiments of the present invention have been described above to facilitate the understanding of the present invention by those skilled in the art, it should be understood that the present invention is not limited to the scope of the embodiments, and various changes may be made apparent to those skilled in the art as long as they are within the spirit and scope of the present invention as defined and defined by the appended claims, and all inventions utilizing the inventive concept are protected.