CN103604411A - Automatic theodolite collimation method based on image recognition - Google Patents

Abstract

The invention discloses an automatic theodolite collimation measurement method based on image recognition. A measurement device of the method comprises an electronic theodolite internally provided with a drive motor, a miniature measurement camera and a fixed fixture, wherein the automatic collimation measurement of the electronic theodolite is guided by the miniature measurement camera, a transformational relation between an image plane coordinate system and a theodolite eyepiece reticule observation coordinate system and a relation between an electronic theodolite deflection angle quantity and a miniature measurement camera pixel quantity when a focal length is in a collimation observation state are calibrated, and collimation return light and an electronic theodolite eyepiece reticule and a deviation relation are automatically extracted until collimation is achieved. According to the automatic theodolite collimation measurement method adopted by the invention, an image recording and analyzing method is used for replacing a conventional human eye observation method, so that the measurement stability in the case of long-term measurement is ensured, and the measurement work efficiency is increased.

Description

Automatic Alignment Method of Theodolite Based on Image Recognition

technical field

The invention belongs to the field of industrial measurement, and in particular relates to a method for automatic alignment measurement of a theodolite.

Background technique

Accuracy measurement is an important guarantee for the reliable operation of satellites and on-board instruments. In order to ensure the normal flight and work of the satellite, it is necessary to measure the geometric accuracy of the satellite's structural shape and precision-required instruments and equipment during ground assembly, such as the attitude of the earth sensor, sun and star sensor, and inertial devices on the satellite. Instruments; thrust components such as 10N thrusters and 490N engines; sensing and communication equipment such as antennas and cameras. The equipment to be measured is generally characterized by a calibrated optical cube mirror on the equipment under test, and the coordinate axis of the coordinate system of the equipment under test is represented by the normal line of the reflective surface of the optical cube mirror. The current measurement method is to use high-precision electronic theodolite for manual alignment measurement. The theodolite has a built-in cross light source that coincides with the eyepiece observation crosshair. When measuring, firstly focus the theodolite telescope lens barrel to infinity, use the cross light source to illuminate the measured cubic mirror surface and observe the return light of the mirror surface through the eyepiece, and operate the theodolite to make the return light The light coincides with the reticle of the eyepiece. At this time, the observation optical axis of the theodolite (irradiation direction of the cross light source) is consistent with the normal direction of the mirror surface. The measurement of the normal direction of the mirror surface is completed through the data recording of the horizontal and vertical angles of the theodolite.

The real-time, non-contact, mobility and high precision of high-precision electronic theodolite measurement technology have significantly improved the speed and accuracy of spacecraft precision measurement, and provided important equipment support for the successful completion of spacecraft precision measurement tasks. However, the current measurement methods mainly rely on manual observation and operation, and the function of the built-in drive motor has not been effectively utilized. In the case of long-term, long-distance, and multi-tasking, the measurement error will change due to the observation error of the human eye. At the same time, the measurement efficiency will be reduced. Especially with the continuous increase of model development tasks, there are often tasks with a small number of people and a large workload, and the observation error of the human eye has gradually become an important link that affects the measurement accuracy of the system. For example, after a worker measures continuously for more than 4 hours, the observation error of the human eye will become larger, especially when working overtime at night, the worker's vision will decrease due to visual fatigue, and the aiming accuracy of the human eye will be lower. It will cause the reduction of measurement efficiency and measurement accuracy, and even delay the progress of the task.

In recent years, due to the advantages of fast detection speed, high measurement accuracy, and non-contact measurement, image sensor technology has shown very strong development prospects in realizing on-site detection, improving production automation, and realizing super-sized detection. Vision-guided measurement technology can overcome the observation errors of the human eye and improve work efficiency to a large extent. It has become a hot spot and focus of research and development in the field of measurement at home and abroad.

However, none of these existing control methods can achieve follow-up control, which is not enough to truly reflect the assembly requirements and perform real-time control.

Contents of the invention

The automatic theodolite alignment measurement method based on image recognition of the present invention is to monitor the alignment measurement state of theodolite through visualized images, utilize the quantifiable analysis characteristics of the image to guide theodolite to perform automatic alignment, and finally realize the automatic alignment of theodolite Measurement, while improving measurement efficiency, eliminates observation errors of the human eye, ensuring or even improving the measurement accuracy of the theodolite.

Technical scheme of the present invention is as follows:

The automatic theodolite collimation measurement method based on image recognition, its measuring equipment is composed of an electronic theodolite with a built-in drive motor, a miniature measuring camera and a fixed tooling, and the electronic theodolite is guided by the miniature measuring camera for automatic collimation measurement, including the following steps:

(1) The fixed tooling connects and fixes the miniature measuring camera with the electronic theodolite, and the miniature measuring camera can observe the measuring field of view of the electronic theodolite through the lens barrel of the electronic theodolite without obstruction;

(2) Calibrate the spatial relationship between the micro-measurement camera and the theodolite as the internal parameters of the system, including the transformation relationship between the image plane coordinate system of the image captured by the camera and the observation coordinate system of the theodolite eyepiece reticle (deflection coefficient and translation amount) and When the focal length is in the state of infinity measurement (the focal length state when the theodolite is collimated), the proportional coefficient between the angle deviation of the electronic theodolite and the pixel deviation in the image captured by the micro-measurement camera (horizontal crosshair direction proportional coefficient and vertical direction proportional coefficient);

a. The conversion relationship (deflection coefficient and translation amount) between the image plane coordinate system of the image captured by the camera and the observation coordinate system of the theodolite eyepiece reticle is as follows:

Let the coordinate system XOY be the reference coordinate system represented by the reticle of the theodolite eyepiece, xoy be the image plane coordinate system of the guiding camera, and the two coordinate systems are linear transformations of rotation and translation of the same plane and the same proportion, and the conversion relationship is set as

X=p×x+q，Y=p×y+q （1）

In the formula (X, Y) is the coordinate value of the target point in the reference coordinate system. The reference coordinate system extracts the theodolite crosshair in the image, and calculates the distance between the target point and the two coordinate axes (measured by the number of pixels) as This coordinate value; (x, y) is the coordinate value of the same target point in the image plane coordinate system. The image plane coordinate system takes the center of the image as the coordinate origin, and the parallel pixel arrangement direction and the vertical pixel arrangement direction constitute the coordinate axis. Calculate the target The distance from the point to the two coordinate axes (measured by the number of pixels) is the coordinate value; p and q are the conversion relationship coefficients;

Take two known points A and B arbitrarily in the image, let the coordinate value of A in the reference coordinate system be (A1, A2), the coordinate value in the image plane coordinate system be (a1, a2), and B be in the reference coordinate system The coordinate values in the index system are (B1, B2), and the coordinate values in the image plane coordinate system are (b1, b2). Through these four sets of coordinate values, the conversion relationship between the reference coordinate system and the image plane coordinate system can be obtained ;

b. The calibration method of the relationship between the deflection angle of the electronic theodolite and the number of pixels of the miniature measuring camera when the focal length is in the collimated observation state is:

Stabilize the electronic theodolite fixedly connected to the miniature measuring camera in state 1, record the horizontal angle and vertical angle of the theodolite under the condition of state 1, record the image of the miniature measuring camera under the condition of state 1, and carry out the horizontal angle and vertical angle of the electronic theodolite Deflection, stable in state 2, record the horizontal angle and vertical angle of the theodolite under state 2 conditions, record the image of the miniature measuring camera under state 2 conditions;

Calculate the pixel amount of image offset in state 1 and state 2, and convert this offset into two offset components in the electronic theodolite observation coordinate system through the calibration relationship in (1), corresponding to the electronic theodolite state 1 and the horizontal angle of state 2 and the deflection of vertical angle, calculate respectively the deflection angle amount of the electronic theodolite on the two coordinate axis directions under the electronic theodolite observation coordinate system and the relational parameter of the miniature measurement camera pixel deviation amount; The position of the different deflection angles of the collimation axis of the theodolite, the relationship between the deflection angle of the electronic theodolite and the pixel deviation of the micro-measurement camera is not completely linear. The parameter calibration method based on the image grid distribution is to divide the image into N×M For several areas, perform a calibration for each area according to the method in step (2). When this parameter needs to be used, select the parameters of the corresponding area based on the area where the target is located in the image according to the principle of proximity;

(3) Automatically extract the relationship between the collimated return light and the electronic theodolite eyepiece reticle and deviation

Using the shape features of the collimated return light and the electronic theodolite eyepiece reticle in the image, the center coordinates of the cross return light and the coordinates of the electronic theodolite eyepiece reticle are extracted in the image through the method based on grayscale analysis, and the two are calculated in the image The pixel deviation amount; the pixel deviation amount is converted into the angular deviation amount of the horizontal axis and the vertical axis in the electronic theodolite observation coordinate system through the relationship parameter calibrated in step (2);

(4) Send the angle deviation calculated in (3) to the electronic theodolite in the form of a command, and drive the theodolite to deflect according to the set angle until the alignment is completed.

In the above scheme, the electronic theodolite used is an electronic theodolite with self-collimation function and a built-in drive motor, and mature electronic theodolite products can be directly used;

In the above scheme, the miniature measurement camera used is a miniature fixed-focus measurement camera, and mature measurement camera products can be directly used;

In the above scheme, the fixed tooling used is a tooling that matches the size of the eyepiece and the miniature measuring camera, which can fix the miniature measuring camera at the position of the eyepiece of the theodolite, and at the same time ensure that the optical axis of the miniature measuring camera and the collimation axis of the theodolite are fixed in parallel. Basically the same.

The automatic collimation method of the theodolite based on image recognition adopted by the present invention replaces the traditional human eye observation method with the method of image recording and analysis, which ensures the measurement stability under long-term measurement conditions and improves the working efficiency of measurement.

Description of drawings

Fig. 1 is a schematic diagram of calibration of the conversion relationship between the image plane coordinate system of the image taken by the camera and the theodolite eyepiece reticle observation coordinate system in the process of implementing the automatic collimation method of the present invention. The XOY indicated by the solid line is the reference coordinate system represented by the reticle of the eyepiece of the theodolite, and the xoy indicated by the dotted line is the image plane coordinate system determined by the image plane.

Fig. 2 is the parameter calibration method based on the image grid distribution in the process of implementing the automatic collimation method of the present invention, which is to divide the image into several regions of N×M, and calibrate each grid point according to the focal length in the collimation observation state . The white point is the center point of the image, and the black point is the grid point of the image.

Fig. 3 is an overall schematic diagram of the theodolite automatic collimation measurement system based on image recognition.

Among them, 1 is theodolite eyepiece; 2 is fixed tooling; 3 is a miniature measuring camera.

Detailed ways

The following is a specific implementation manner of the content of the present invention, and the content of the present invention will be further clarified through the specific implementation mode below. Of course, the following specific embodiments are described only to illustrate different aspects of the present invention, and should not be construed as limiting the scope of the present invention.

As shown in Figure 3, the equipment that the automatic theodolite collimation measurement method based on image recognition of the present invention uses comprises miniature measuring camera, and this miniature measuring camera is fixed on the eyepiece place of theodolite by fixed frock, between miniature measuring camera and fixed frock and The fixing tool and the eyepiece are fixed by screw clamping. The field of view of the electronic theodolite is observed through the measuring lens barrel of the electronic theodolite, and two kinds of calibrations are carried out successively during the measurement. The specific implementation methods are as follows:

(1) The conversion relationship (deflection coefficient and translation amount) between the image plane coordinate system of the image captured by the camera and the observation coordinate system of the theodolite eyepiece reticle is as follows:

Let p and q be the conversion relation coefficients. As shown in Figure 1, it is the image taken by the camera. Use the manual drawing method to select two points on the crosshair with the mouse to obtain the horizontal crosshair line and the vertical crosshair line, and record the center of the crosshair and the point on the horizontal crosshair and the vertical crosshair. The pixel position of the point on the crosshair in the image. The center point of the image is o, the row direction of the pixel arrangement is ox, the column direction of the pixel arrangement is oy, the center of the theodolite eyepiece crosshair is displayed as O in the image, the direction of the horizontal crosshair is OX, and the direction of the vertical crosshair is OY. Let the coordinate system XOY be the reference coordinate system represented by the crosshairs, and xoy be the image plane coordinate system of the guiding camera. The two coordinate systems are rotation and translation in the same plane, and the mathematical relationship is a linear relationship. Let this conversion relationship be

X=p×x+q，Y=p×y+q （1）

In the formula (X, Y) is the pixel coordinate value of the target point in the reference coordinate system. The reference coordinate system extracts the theodolite crosshair in the image, and calculates the distance from the target point to the two coordinate axes (measured by the number of pixels). is this coordinate value; (x, y) is the pixel coordinate value of the same target point in the image plane coordinate system, the image plane coordinate system takes the image center as the coordinate origin, and the parallel pixel arrangement direction and the vertical pixel arrangement direction constitute the coordinate axis, This coordinate value is used to calculate the distance from the target point to the two coordinate axes (measured in pixels).

Take point A arbitrarily in the image, set the coordinate value of A in the reference coordinate system as (A1, A2), and the coordinate value in the image plane coordinate system as (a1, a2). Then there is the equation;

A1=p×a1+q and A2=p×a2+q;

Solve the equation to get: p=(A1-A2)/(a1-a2), q=A1-(A1-A2)×/(a1-a2).

(2) When the focal length is in the state of infinity measurement (the focal length state when the theodolite is collimating), the calibration method of the proportional coefficient between the angle offset of the electronic theodolite and the pixel offset in the image captured by the micro-measurement camera is as follows:

Let K be the scale factor in the horizontal direction, and K' be the scale factor in the vertical direction. There are two points C and D in the image. The pixel coordinate value of point C in the xoy coordinate system is (C, D), and the pixel coordinate value of point D in the xoy coordinate system is (c, d). After passing through the coordinate system The conversion relationship between XOY and xoy can get the pixel coordinate value of point C in the XOY coordinate system (C', D'), and the pixel coordinate value of point D in the XOY coordinate system (c', d'). When the theodolite is aimed at point C, the angle of the theodolite is displayed as (α (horizontal angle), β (vertical angle)), when the theodolite is aimed at point D, the angle of theodolite is displayed as (α' (horizontal angle), β' (vertical angle) horn)).

Then the proportional coefficient K in the horizontal direction is: K=|α′-α|/|C′-C|; (if the theodolite turns over the zero point of the horizontal angle when the theodolite is aimed at point D, then K=|α′+360- α|/|C′-C|;)

The proportional coefficient K' in the vertical direction is: K'=|β'-β|/|D'-D|.

Since the relationship between the deflection angle of the electronic theodolite and the pixel deviation of the micro-measurement camera at different deflection angles from the collimation axis of the theodolite is not completely linear, a parameter calibration method based on image grid distribution is adopted, as shown in Figure 2. Divide the image into several N×M areas according to the row and column direction of the image, where the white origin is the center of the image, and define other row and column intersection points (black points) as (D1, D2...) and select it when performing (2) ② calibration Image center point C, sequentially select (D1, D2...) as point D for calibration, and obtain the proportional coefficient K in the horizontal direction and the proportional coefficient K' in the vertical direction at each point D1, D2... .

(3) In the measurement state, the collimated measurement field of view of the theodolite is recorded by the miniature measuring camera, and the collimated cross light in the image is segmented by the Steger method in the recorded image, and the collimated return light of the electronic theodolite is extracted The horizontal centerline point and the vertical centerline point, and then collimate and return the light centerline through the straight line extraction method based on Hough transform, and obtain the pixel coordinate value (m,n) of the crosshair center in the image coordinate system xoy through intersection calculation , calculate the row-column intersection point (D1, D2...) closest to the (m, n) point, and set the proportional coefficient in the horizontal direction of the intersection point to K, and the proportional coefficient in the vertical direction to be K′. The internal parameters calibrated in (2) are converted into the angular driving quantities (γ, δ) of the electronic theodolite, and the specific calculation method is:

γ=(m×p+q)×K;

δ=(n×p+q)×K′.

(4) According to the angular drive amount of the electronic theodolite in (3), the computer sends control instructions to drive the electronic theodolite to deflect the angle according to (γ, δ) to achieve the collimation state.

Although the specific embodiments of the present invention have been described and illustrated in detail above, it should be pointed out that we can make various changes and modifications to the above embodiments, but these do not depart from the spirit of the present invention and the appended rights. Request the range described.

Claims (5)

1. the automatic transit alignment measurement method based on image recognition, its measuring equipment is comprised of the electronic theodolite of built-in CD-ROM drive motor, miniature measurement camera and fixing tool, by miniature measurement camera, electronic theodolite is carried out to the guiding of autocollimation measurement, comprise the steps:

(1) fixing tool is connected and fixed miniature measurement camera with electronic theodolite, and miniature measurement camera can carry out unobstructed observation to the measurement visual field of electronic theodolite by electronic theodolite lens barrel;

(2) calibrating miniature is measured the spatial relationship of camera and transit as the work inner parameter of system, comprises that transformational relation (deflection factor and translational movement) between being of the plane of delineation coordinate system of the captured image of camera and transit eyepiece crosshair observation coordinate and focal length survey under state (focal length state when transit collimates) scale-up factor of pixel deviator in electronic theodolite angle deviator and miniature measurement image shot by camera in infinite distance;

A. transformational relation (deflection factor and the translational movement) scaling method between the plane of delineation coordinate system of the captured image of camera and transit eyepiece crosshair observation coordinate system is:

If coordinate system XOY is the frame of reference that transit eyepiece crosshair characterizes, xoy is the photo coordinate system of guiding camera, and two coordinates are isoplanar rotation and translation linear transformation in proportion, establish this transformational relation to be

X=p×x+q，Y=p×y+q （1）

In formula, (X, Y) is the coordinate figure of impact point in the frame of reference, and the frame of reference is extracted by the transit crosshair in image, and the distance (pixel count metering) of calculating impact point to two coordinate axis is this coordinate figure; (x, y) be the coordinate figure of same impact point in photo coordinate system, photo coordinate system be take picture centre as true origin, and parallel pixel orientation and vertical pixel orientation form coordinate axis, and the distance of calculating impact point to two coordinate axis is this coordinate figure; P, q are transformational relation coefficient;

In image, get arbitrarily two known point A, B, if the coordinate figure of A in the frame of reference is (A1, A2), coordinate figure in photo coordinate system is (a1, a2), and the coordinate figure of B in reference index system is (B1, B2), coordinate figure in photo coordinate system is (b1, b2), can solve the transformational relation of the frame of reference and photo coordinate system by these four groups of coordinate figures;

B. focal length electronic theodolite deflection angle tolerance under collimation observer state with the scaling method of the relation of miniature measurement camera pixel quantity is:

The electronic theodolite that is fixedly connected with miniature measurement camera is stable to state 1, the level angle of transit and vertical angle under recording status 1 condition, the image of the miniature measurement camera under recording status 1 condition, electronic theodolite is carried out to level angle and vertical angle deflection, be stable at state 2, the level angle of transit and vertical angle under recording status 2 conditions, the image of the miniature measurement camera under recording status 2 conditions;

The amount of pixels of image shift in computing mode 1 and state 2, by the demarcation relation in (1), this side-play amount dress is changed to two offset components under electronic theodolite observation coordinate system, corresponding to the level angle of electronic theodolite state 1 and state 2 and the deviator of vertical angle, calculate respectively the deflection angle tolerance of the electronic theodolite on two change in coordinate axis direction under electronic theodolite observation coordinate system with the Relation Parameters of miniature measurement camera pixel departure; In the position apart from the different deflection angles of transit collimation axis, the deflection angle tolerance of electronic theodolite is not linear relationship completely with the relation of miniature measurement camera pixel departure, the parameter calibration method distributing based on image lattice, be about to some regions that image is divided into N * M, each graticule mesh intersection point is once demarcated according to the method in (b), in the time need to using this parameter, according to target, according to principle nearby, selected the parameter of corresponding grid points position with the distance of possessive case site in image;

(3) automatically extract collimation back light and electronic theodolite eyepiece crosshair and deviation relation

Utilize collimation back light and electronic theodolite eyepiece crosshair shape facility in image, by the method based on gray analysis, in image, extract the centre coordinate of cross back light and the coordinate of electronic theodolite eyepiece crosshair, calculate the two pixel departure in image; By the Relation Parameters of demarcating in (2), pixel departure is converted into the amount of angular deviation of transverse axis and vertical axes under electronic theodolite observation coordinate system;

(4) amount of angular deviation of calculating in (3) is sent to electronic theodolite with instruction type, drive transit to carry out deflection according to set angle, until complete collimation.

2. the method for claim 1, wherein the scale-up factor of pixel deviator comprises planche cross silk direction scale-up factor and vertical direction scale-up factor in electronic theodolite angle deviator and miniature measurement image shot by camera.

3. the method for claim 1, wherein electronic theodolite is the electronic theodolite that has autocollimation function and be built-in with CD-ROM drive motor.

4. the miniature measurement camera the method for claim 1, wherein adopting is the miniature measurement camera that focuses.

5. the method as described in claim 1-4 any one, wherein, the fixing tool adopting is for coordinating the frock of eyepiece and miniature measurement camera size, when miniature measurement camera can be fixed on to transit eyepiece position, by parallel fixing mode guarantee the optical axis of miniature measurement camera and the collimation axis of transit basically identical.

Images