CN113503830B

CN113503830B - Aspheric surface shape measuring method based on multiple cameras

Info

Publication number: CN113503830B
Application number: CN202110756700.3A
Authority: CN
Inventors: 刘旭; 潘银飞
Original assignee: Wuxi Dimension Investment Management Partnership LP
Current assignee: Wuxi Dimension Investment Management Partnership LP
Priority date: 2021-07-05
Filing date: 2021-07-05
Publication date: 2023-01-03
Anticipated expiration: 2041-07-05
Also published as: CN113503830A

Abstract

The invention discloses a multi-camera-based aspheric surface shape measuring method, which is applied to a system consisting of a reference plane, a camera, an optical platform and a computer, and comprises the following measuring steps: firstly, completing the calibration of the position of each camera and a reference plane; then, obtaining the relative position relation among the cameras by utilizing the identity of the reference plane position; and finally, a camera acquires a coded image modulated by the curvature change of the surface of the high-reflection object to be measured, and the coded image information is demodulated to recover the three-dimensional information of the surface of the object to be measured, so that the overall measurement of the surface of the object to be measured is realized through the relative position conversion relation of multiple cameras.

Description

Aspheric surface shape measuring method based on multiple cameras

Technical Field

The invention relates to a measuring method based on multiple cameras, in particular to a method for measuring the surface shape of a high-reflection aspheric surface.

Background

The optical detection technology based on machine vision is used as a non-contact measurement means, and has the advantages of high measurement speed, moderate precision and no damage to the surface of an object to be measured. Since the last 70 s, it has been considered one of the most promising technologies for solving measurement problems in industrial manufacturing. However, the existing mature vision measurement systems are only suitable for measuring objects with diffuse reflection surfaces, and when aspheric surfaces with high reflection properties are measured, large errors can be caused or measurement cannot be performed. And due to the influence of the curvature of the surface to be measured, only local measurement of the surface to be measured can be realized by a single camera.

Disclosure of Invention

In order to overcome the defects in the prior art, the invention provides the aspheric surface shape measuring method based on the multiple cameras, so that the problem of locality of aspheric surface shape measurement by a single camera can be solved, and the overall measurement of the aspheric surface shape is realized.

In order to achieve the purpose, the invention adopts the following technical scheme:

the invention relates to a multi-camera-based aspheric surface shape measuring method which is characterized by being applied to a measuring system consisting of a first camera, a second camera, a third camera, a fourth camera, a fifth camera, a computer, an object to be measured, an optical platform and a reference plane, wherein the object to be measured is placed on the optical platform, the reference plane is arranged above the optical platform, and a camera is respectively placed at the four vertex angles and the central position of the reference plane; each camera is respectively connected with the computer;

the aspheric surface shape measuring method comprises the following steps:

step 1, calibrating the relative positions of the five cameras and a reference plane respectively to obtain a position conversion relation between a camera coordinate system and a reference plane coordinate system as shown in a formula (1):

X _i ＝R _i ·P _w +T _i (1)

in the formula (1), R _i And T _i Respectively representing a rotation matrix and a translation vector from a reference plane coordinate system to an ith camera coordinate system; p _w Representing coordinates of a w-th characteristic corner point on a reference plane; w is E [1,N](ii) a N denotes the number of characteristic corner points in the reference plane, X _i Representing the coordinates of the characteristic corner point in the ith camera coordinate system; i is an E [1,5]；

Step 2, determining the relative position relationship between any two camera coordinate systems by using the formula (2):

in the formula (2), the reaction mixture is,

representing the inverse of a rotation matrix, T, representing the reference plane coordinate system to the j-th camera coordinate system _j Representing a translation vector, X, representing the reference plane coordinate system to the jth camera coordinate system _j Represents the coordinate of the characteristic corner point in the j-th camera coordinate system, i, j is E [1,5 ∈ [ ]]；i≠j；

Obtaining rotation matrixes R from the jth camera coordinate system to the ith camera coordinate system by using the formulas (3) and (4) _ij And translation vector T _ij ：

Step 3, the five cameras respectively collect the coded images of the object to be detected and send the coded images to the computer, and the computer demodulates the coded images to obtain a surface three-dimensional coordinate set of the object to be detected); and recording a surface point three-dimensional coordinate set of the object to be detected, which is obtained from the coded image acquired by the jth camera system, as S _j ；

And (3) making the camera coordinate system at the central position be a reference coordinate system, and obtaining the surface integral three-dimensional information S of the object to be measured by using the formula (5), thereby completing the integral measurement of the aspheric surface shape:

S＝R _ij ·S _j +T _ij (5)。

compared with the prior art, the invention has the beneficial effects that:

1. according to the invention, a multi-camera measuring system is built, the aspheric object to be measured is shot from different angles, and the three-dimensional shape measurement of each part of the aspheric surface shape is carried out, so that the problem of locality of the aspheric surface shape measured by a single camera is solved, and the overall measurement of the aspheric surface shape is completed.

2. The coding pattern and the size on the reference plane can be set by self, and the operation flexibility is strong.

3. The multi-camera non-contact measurement ensures the reliability and the precision of the surface shape detection.

4. The invention can be applied to the measurement of the surface shape of the bright surface of the automobile windshield, the automobile body shell, the spherical mirror and the like, and has good practical applicability.

Drawings

FIG. 1 is a block diagram of the system of the present invention;

reference numbers in the figures: 1 a first camera; 2 a second camera; 3 a third camera; 4 a fourth camera; 5 a fifth camera; 6, a computer; 7, an object to be detected; 8 an optical bench; 9 reference plane.

Detailed Description

In this embodiment, a method for measuring an aspheric surface shape based on multiple cameras is applied to a measurement system composed of a first camera 1, a second camera 2, a third camera 3, a fourth camera 4, a fifth camera 5, a computer 6, an object to be measured 7, an optical platform 8, and a reference plane 9, as shown in fig. 1, where the object to be measured 7 is placed on the optical platform 8, the reference plane 9 is disposed above the optical platform 8, and a camera is placed at each of four vertex angles and a central position of the reference plane 9; each camera is respectively connected with the computer 6; the multi-camera aspheric surface shape measuring method comprises the following steps:

step 1, calibrating the relative positions of the five cameras and the reference plane 9 respectively. Each camera respectively collects mirror images of calibration patterns on a reference plane and transmits the mirror images back to the computer 6, then the reference plane is calibrated by firstly calibrating the mirror images of the reference plane through a Zhang Yongyou calibration method, and then the relative position relation between a camera coordinate system and a real reference plane is obtained by combining the law of mirror reflection, so that the position conversion relation between the camera coordinate system and the reference plane coordinate system shown in the formula (1) is obtained:

X _i ＝R _i ·P _w +T _i (1)

in the formula (1), R _i And T _i Respectively representing a rotation matrix and a translation vector from a reference plane coordinate system to an ith camera coordinate system; p _w The coordinates of the w characteristic corner point on the reference plane (9) are represented; w is E [1,N](ii) a N denotes the number of characteristic corner points in the reference plane (9), X _i Representing the coordinates of the characteristic corner point in the ith camera coordinate system; i is an E [1,5]；

Step 2, determining the relative position relation between any two camera coordinate systems by using the formula (2):

in the formula (2), the reaction mixture is,

representing the inverse of a rotation matrix, T, representing the reference plane coordinate system to the j-th camera coordinate system _j Representing a translation vector, X, representing the reference plane coordinate system to the jth camera coordinate system _j Represents the coordinate of the characteristic corner point in the j-th camera coordinate system, i, j is E [1,5 ∈ [ ]]；i≠j。

Step 3, the five cameras respectively collect coded images of the object (7) to be detected and send the coded images to the computer (6), and the computer (6) demodulates the coded images to obtain a surface three-dimensional coordinate set of the object (7) to be detected; when the relative position of a reference plane is known, a coding pattern on the reference plane is imaged on an image plane through a high-reflection surface of an object to be detected, dense reflection correspondence among two-dimensional characteristic points of a normalized image plane, three-dimensional characteristic points of the reference plane and surface reflection points of the object to be detected is established, the depth of the reflection points of the high-reflection surface is parameterized, a curved surface where the reflection points are located is fitted through a polynomial, and the depth of the reflection points is iteratively calculated by using an LM optimization method in combination with the uniqueness and second-order continuity of a normal vector of the high-reflection surface. Recording the surface point three-dimensional coordinate set recovered by the jth camera system as S _j And enabling the camera coordinate system at the central position to be a reference coordinate system, and obtaining the surface integral three-dimensional information S of the object to be measured (7) by using the formula (5), thereby completing the integral measurement of the aspheric surface shape:

S＝R _ij ·S _j +T _ij (5)。

Claims

1. a multi-camera-based aspheric surface shape measurement method is characterized by being applied to a measurement system consisting of a first camera (1), a second camera (2), a third camera (3), a fourth camera (4), a fifth camera (5), a computer (6), an object to be measured (7), an optical platform (8) and a reference plane (9), wherein the object to be measured (7) is placed on the optical platform (8), the reference plane (9) is arranged above the optical platform (8), and a camera is respectively placed at four top angles and the central position of the reference plane (9); each camera is respectively connected with a computer (6);

the aspheric surface shape measuring method comprises the following steps:

step 1, calibrating the relative positions of the five cameras and a reference plane (9) respectively to obtain a position conversion relation between a camera coordinate system and a reference plane coordinate system as shown in a formula (1):

X _i ＝R _i ·P _w +T _i (1)

in the formula (2), the reaction mixture is,

an inverse matrix, T, representing a rotation matrix of the reference plane coordinate system to the j-th camera coordinate system _j Representing a translation vector, X, from the reference plane coordinate system to the jth camera coordinate system _j Represents the coordinate of the characteristic corner point in the j-th camera coordinate system, i, j is E [1,5 ∈ [ ]]；i≠j；

Step 3, the five cameras respectively collect the coded images of the object (7) to be detected and send the coded images to the computer (6), and the computer (6) demodulates the coded images to obtain a surface three-dimensional coordinate set of the object (7) to be detected; wherein, the three-dimensional coordinate set of the surface point of the object (7) to be measured obtained by the coded image collected by the jth camera system is recorded as S _j ；

And (3) making the camera coordinate system at the central position be a reference coordinate system, and obtaining the surface integral three-dimensional information S of the object to be measured (7) by using the formula (5), thereby completing the integral measurement of the aspheric surface shape:

S＝R _ij ·S _j +T _ij (5)。