CN102116607B - Method and device for measuring axial displacement characterized by one-dimensional (1D) contrast ratio - Google Patents

Abstract

A method and device for measuring axial displacement characterized by one-dimensional contrast is composed of an ordinary computer connected to a computer camera, a stepping motor and its interface circuit, and a camera axial displacement device through its USB interface, and the camera is installed On the axial displacement device, it is driven by a stepping motor, and the stepping motor is connected to the RS232C interface of the computer through the stepping motor interface circuit. The computer is equipped with a camera to shoot and measure the axial displacement according to the one-dimensional contrast program. The patent of the present invention uses one-dimensional contrast as the feature of the image frame of the measured object, automatically analyzes and selects the best observation area by calculating the self-correlation matching coefficient of pixel brightness; and judges the imaging of the object by counting the number of image features in the observation window area The degree of focus, and then measure the tiny displacement of the object in the direction of the optical axis of the camera; this measurement method is novel and can adapt to a certain degree of environmental light changes.

Description

Method and device for measuring axial displacement characterized by one-dimensional contrast

technical field

The invention belongs to the technical field of digital image measurement, in particular to a method and a device for measuring the tiny displacement of an object along its optical axis using a computer camera.

Background technique

The recently filed invention patent application "Method and device for measuring displacement along the optical axis direction based on the contrast ratio of three primary colors" analyzes the technology for detecting the displacement occurring along the optical axis direction of the system, and proposes a novel method for detecting axial displacement The method and the device thereof make full use of the information contained in the image frame captured by the camera, however, the analysis calculation amount is relatively large, which affects the measurement speed.

Contents of the invention

The invention provides a method and device for measuring axial displacement characterized by one-dimensional contrast. It uses a computer camera to measure the tiny displacement of an object along the optical axis of the camera in an environment where the lighting conditions change to a certain extent. vector.

The technical solution adopted by the present invention to solve its technical problems is: an ordinary computer is equipped with a computer camera, and the camera is installed on an axial displacement device composed of a high-precision micro-displacement stepping motor as the core. The motor is connected to the RS232C interface of the computer through a stepper motor interface circuit, and the computer is equipped with a program for shooting with a camera and measuring the axial displacement according to the one-dimensional contrast ratio, which reflects the characteristic measurement axial displacement with the one-dimensional contrast ratio as the image frame. Displacement methods, including:

Step 1. Take a frame of an image of the measured object in the format of a bitmap (M×N, M, N∈ positive integer) as a reference frame; take the position of the first pixel in the upper left corner of the pixel array of the frame as the origin , taking the rightward direction as the x-axis direction, and the vertical downward direction as the y-axis direction, the unit of the coordinate system is the size of one pixel; select an area in the central area of the pixel array, and the size is m ₀ ×n ₀ , m ₀ , n ₀ ∈ a positive integer, called the observation window, which is h and v pixels away from the edge pixels in the horizontal direction and vertical direction of the pixel array, that is: m ₀ +2h=M, n ₀ +2v=N, h, v∈ positive integer;

Step 2. For the pixel array of the above reference frame, derive the side direction data along the X-axis direction and the Y-axis direction pixel by row and pixel by column, and use 3-bit binary values 001, 010 and 100 to represent the positive side, The negative edge and the third type edge constitute two frames of edge direction data {reference _x (x, y)} and {reference _y (x, x, y)}, wherein, the subscript x or y respectively represent the direction of the coordinate axis along, and the symbol "{ }" represents the pixels (x, y) in the observation window along the direction of the coordinate axis indicated by the function subscript A collection of edge direction data, save these data;

Step 3. For the above two frames of edge direction data, calculate the self-correlation matching coefficient of the pixel array in the observation window in the reference frame:

$\mathrm{<span\; class="notranslate">\; auto</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; correlation</span>}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

$\mathrm{<span\; class="notranslate">\; auto</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; correlation</span>}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

In the formula, the operation symbol · represents the binary logic AND operation, and the operation result is either logic 0 or logic 1, and the operation symbol "[ ]" represents the value corresponding to the value of the logic operation function, or the value 0, or The value is 1, and the combination of parameter variables a and b determines the scale of the associative matching operator array. If a 3×3 associative matching operator array is used: a=-1, 0, 1, b=-1, 0, 1, Therefore, 9 auto-correlation coefficients auto_correlation _x (a, b) and auto_correlation _y (a, b) will be generated along each axis direction;

Step 4. According to the self-association matching coefficients corresponding to the above two frames of edge direction data, search for the best observation window pixel array that can be matched and compared under the current object surface conditions and lighting conditions:

m _x = m ₀ ± step, n _x = n ₀ ± step, 2h = Mm _x , 2v = Nn _x

and m _y = m ₀ ± step, n _y = n ₀ ± step, 2h = Mm _y , 2v = Nn _y ,

In the formula, the subscripts x and y indicate that their values correspond to the directions along the X-axis and the Y-axis respectively, step is the step parameter in the search process, and the plus and minus signs in front of it are determined by the search direction; take the two sets of values The larger one is the size of the observation window array used for this measurement: m×n;

Step 5. For the pixel array of the above-mentioned reference frame, according to the brightness of the pixels, the data along the X-axis direction of one frame edge direction is derived row by row;

According to the above edge direction data, for the observation window area therein, use accumulators to count the numbers of their corresponding positive and negative sides: N _{X-axis positive} and N _{X-axis negative} , where N (umber) represents the number; accumulate The counting result of the above-mentioned accumulator and save it, expressed as: N(i, j=0, forw=0, back=0), wherein, i=1, 2, 3... represents the taken reference The sequence count of the frame is also the count of the measurement, j=0, 1, 2, 3, ..., represents the count of the sampling frame taken during the i-th measurement process, the variable forw(=0, 1, 2 ,...) and back(=0, 1, 2,...) represent the steps corresponding to the clockwise rotation and counterclockwise rotation of the stepping motor in the i-th measurement, respectively Pulse counting, before the start of this measurement, there are: i=1, j=0, forw=0, back=0;

Step 6, measurement start: the computer outputs the first forw=1 digital pulse signal to the stepper motor interface circuit through an output control line FORWARD of its RS232C interface, and controls the stepper motor to rotate one step clockwise, and then, Take the sample frame bitmap of j=1 frame;

For the pixel array of the above sampling frame, according to its pixel brightness, export its side direction data of one frame along the X-axis direction line by line;

According to the above side direction data, for the observation window area, use the accumulator to count the number of their corresponding positive side and negative side: N _{X-axis positive} and N _{X-axis negative} ; accumulate the counting results of the above accumulator and save it , expressed as: N(i, j=1, forw=1, back=0);

Step 7, if: N (i, j=1, forw=1, back=0) ≥ N (i, j=0, forw=0, back=0), described computer is through one output of its RS232C interface The control line FORWARD outputs the forw (=2, 3, ...) digital pulse signal to the stepping motor interface circuit, controls the stepping motor to rotate clockwise and advances, and takes a picture every time clockwise rotates one step And analyze and compare the sum N (i, j=forw-1, forw-1, back=0) and N (i, j =forw, forw, back=0), save it, there should be in this process: N(i, j=forw, forw, back=0)≥N(i, j=forw-1, forw-1, back =0), until satisfying: N(i, j=forw, forw, back=0)<N(i, j=forw-1, forw-1, back=0), at this moment, record (forw) _MAX = Forw, at the same time, the computer outputs a digital pulse signal (back=1) to the stepper motor interface circuit through another output control line BACKWARD of its RS232C interface to control the stepper motor to rotate one step counterclockwise, at this time , relative to the initial position before the start of this measurement, the camera of this device is in the best object imaging focus position in this measurement: FocusP=(forw) _MAX -back=(forw) _MAX -1, the calculated result is greater than Or equal to 0, indicating that the direction of rotation of the stepper motor is clockwise, and the sum of the numbers of all positive and negative sides along the X-axis direction in the observation window of the sampling frame corresponding to the focus position is: N(i, j=(forw) _MAX -1, forw=(forw) _MAX -1, back=0), the total counting result of the sample frames taken during the measurement is: j=(forw) _MAX +back=(forw) _MAX + 1;

If: N(i, j=1, forw=1, back=0)<N(i, j=0, forw=0, back=0), said computer is through an output control line BACKWARD of its RS232C interface Output the first back=1 digital pulse signal to the stepper motor interface circuit, control the stepper motor to take a step back in counterclockwise rotation, that is, return to the initial position of this measurement, and then, the computer outputs through its RS232C interface The control line BACKWARD continues to output the back (=2, 3, ...) digital pulse signal to the stepper motor interface circuit to control the stepper motor to further counter-clockwise rotation. In the process, each time One step of anti-clockwise rotation is all taken and analyzed and compared before and after each sampling frame observation window along the X-axis direction with the sum of the numbers N(i, j=back+1, forw=1, back) and N( i, j=back, forw=1, back=back-1), there should be: N(i, j=back+1, forw=1, back=back)≥N(i, j=back, forw=1 , back=back-1), save it, and continue like this until it is satisfied: N(i, j=back+1, forw=1, back)<N(i, j=back, forw=1, back-1) , at this moment, record: (back) _MAX =back, simultaneously, described computer outputs a digital pulse signal (forw=2) to described stepping motor interface circuit by another root output control line FORWARD of its RS232C interface, controls The stepper motor rotates one step clockwise. At this time, relative to the initial position before the start of this measurement, the camera of this device is at the best object imaging focus position in this measurement: FocusP=forw-(back) _MAX =2 -(back) _MAX , the calculation result is a negative value, indicating that the rotation direction of the stepping motor is counterclockwise, and all positive and negative sides along the X-axis direction in the observation window of the sampling frame corresponding to the focus position The sum of the numbers is: N(i, j=((back) _MAX -1)+1=(back) _MAX , forw=1, back=(back) _MAX -1), the sampling frame taken during the measurement Total counting result is: j=forw+(back) _MAX =2+(back) _MAX ;

To sum up, relative to the initial position before the start of this measurement, the displacement of the object obtained in this measurement along the optical axis direction of the camera is: Δz(i)=forw-back, the counters forw and back in the above calculation formula The values of are all the final counting results in this measurement process, and the calculation results have positive or negative signs, which respectively represent the direction in which the displacement occurs along the optical axis to advance or retreat, corresponding to the direction of rotation of the stepping motor (ie Clockwise or counterclockwise), specifically determined by the arrangement of the axial displacement device of the stepper motor;

The total displacement is: ΔZ ₀ (i)=ΔZ ₀ (i-1)+Δz(i)

Among them, ΔZ ₀ (i-1) is the accumulated axial displacement before this measurement;

Step 8. Prepare for the next measurement: measure the number of times counter i=i+1, take all positive sides and negative sides along the X-axis direction in the corresponding object imaging frame observation window at the best object imaging focus position determined by the ith measurement The sum of the number of sides is used as the new measurement reference value:

N(i, j=0, forw=0, back=0)=N(i, j=(forw) _MAX -1, forw=(forw) _MAX -1, back=0), or N(i, j =0, forw=0, back=0)=

N(i, j=((back) _MAX -1)+1=(back) _MAX , forw=1, back=(back) _MAX -1);

Step 9, jump to step 6, continue to measure;

In the actual measurement process, direct measurement results can be obtained through measurement calibration.

The definition of the side direction data in the above camera shooting and the step of measuring the axial displacement program according to the one-dimensional contrast ratio is:

In the pixel array, along the X-axis or along the Y-axis, if the light intensity value of a pixel is smaller than the corresponding light intensity value of the second pixel behind it by an error tolerance value error, that is, if

I(X,Y)<I(X+2,Y)-error or I(X,Y)<I(X,Y+2)-error

It is defined that there is a positive edge along the axis between these two pixels; if the light intensity value of a pixel is greater than the corresponding light intensity value of the second pixel behind it by an error tolerance value error, that is, if

I(X,Y)>I(X+2,Y)+error or I(X,Y)>I(X,Y+2)+error

Then it is defined that there is a negative edge along the axis between these two pixels; the edge obtained in this way is located at the position of the first pixel after this pixel, that is, the pixel located in the middle of the two pixels involved in the comparison ; If a certain light intensity value of a pixel is close to the corresponding light intensity value of the second pixel behind it, the value difference does not exceed an error tolerance value error, that is, if

I(X+2, Y)-error≤I(X, Y)≤I(X+2, Y)+error

or I(X, Y+2)-error ≤ I(X, Y) ≤ I(X, Y+2)+error,

It is considered that there is no corresponding "edge" along the axis between the two pixels, or it is called the third type of edge;

Along a certain coordinate axis direction, all positive edges, negative edges, and third-type edges of the corresponding pixel row or pixel column form the edge direction data of the row or column along the coordinate axis direction; the error tolerance in the above formula The limit value can be preset as a small value according to the specific lighting conditions, for example: error=10; there is no edge direction data at the pixel positions on the four sides and corners in the pixel array.

The method of searching for the optimal observation window pixel array according to the steps of the camera shooting and measuring the axial displacement according to the one-dimensional contrast ratio includes:

For the observation window of the pixel array and the k×k (k ∈ positive integer) associative matching operator array (a, b), k×k self-association matching coefficients will be generated along a certain coordinate axis, and compared according to the following inequality These self-association matching coefficients:

auto_correlation(a,b)≥auto_correlation(0,0)×similarity

In the formula, similarity describes the similarity between the observation window and its adjacent pixel array of the same scale, for example, similarity = 60%, which can be set in advance, or can be debugged and selected according to the lighting conditions and the texture of the surface of the measured object;

If there are more than k×k×1/3 self-correlation coefficients satisfying the above inequality, it is necessary to expand the scope of the observation window for each step row and step column: let m=m ₀ +step, n=n ₀ +step, and recalculate the new The self-correlation coefficients of the observation window, and carry out the above comparison until the self-correlation matching coefficients satisfying the above inequality are no more than k×k×1/3, at this time, 2h=Mm, 2v=Nn, where step is the step The input parameter, the initial value is 1, every time the scale of the observation window needs to be expanded, it will increase by 1; if it exceeds a predetermined range in the frame, and no suitable observation window has been found, it is considered that the texture of the reflective surface of the object is not suitable for this part. The measurement work of the device, and give prompts and warnings;

If there are no more than k×k×1/3 self-correlation matching coefficients satisfying the above inequality, it means that the surface structure features of the object to be photographed are fine enough, and the values between the nearest adjacent pixels can be distinguished, and further attempts can be made to reduce the size of the observation window. Range each step row and step column to reduce the calculation workload: let m=m ₀ -step, n=n ₀ -step, recalculate the self-correlation coefficient of the observation window, and perform the above comparison, and the progressive parameter step increases each time 1. Until the number of self-correlation coefficients satisfying the above inequality in the selected observation window area is not less than k×k×1/3, at this time, it is considered that the optimal observation window pixel array has been searched.

The camera axial displacement device includes: the camera is installed on a worktable, and the workbench is connected with a long threaded rotating shaft with threaded thread, and the threaded rotating shaft is installed on a On the large workbench, and it is socketed with the two support frames in the form of a rotating shaft, and can rotate at the socket with the support frame but does not move forward or backward; the screw thread rotates There is a fixed gear on the shaft, which meshes with the gear on the shaft of the stepper motor. The stepper motor is also installed on the large workbench, and it is connected to the RS232C interface of the computer system through the stepper motor interface circuit.

The patent of the invention uses the one-dimensional contrast of pixel brightness as the feature of the image frame of the measured object, automatically analyzes and selects the best observation area by calculating the self-correlation matching coefficient; and judges the imaging focus of the object by counting the number of image features in the observation area degree, and then measure the tiny displacement of the object in the direction of the optical axis of the camera; this measurement method is novel, and compared with the "method and device for measuring displacement along the optical axis with the contrast of three primary colors", the advantages of the patent of the present invention are, The analysis operation load is reduced, and the measurement speed is improved.

Description of drawings

Further illustrate the patent of the present invention below in conjunction with accompanying drawing.

Fig. 1 is a block diagram of a computer and a camera measurement system thereof of the present invention.

Fig. 2 is a block diagram of the camera axial displacement device composed of a high-precision micro-displacement stepping motor as the core of the present invention.

Fig. 3 is a schematic diagram of the optical imaging focusing process of the present invention.

Fig. 4 is a schematic diagram of a pixel array and its observation window area generated after the photoelectric sensor chip undergoes photoelectric conversion.

Fig. 5 is a schematic diagram of a row of optical signals and their digitized signals, edge direction data.

In Figure 1, 1. Computer camera, 2. Optical lens, 3. Photoelectric sensor chip, 4. USB interface, 5. Computer system, 6. USB interface, 7. CPU, 8. RS232C interface, 9. Display card and monitor , 10. Memory and hard disk, 11. Keyboard and mouse, 12. Operating system, 13. Camera driver, 14. Camera shooting and axial displacement measurement program based on one-dimensional contrast ratio, 15. Lighting equipment.

Among Fig. 2, 30. the workbench of stepper motor and camera, 31. bracing frame, 32. bracing frame, 33. the workbench of camera (1), 34. thread thread rotating shaft, 341. thread thread rotating shaft (34 ), the gear on the 40. stepping motor, the rotating shaft of 41. stepping motor (40), the gear on the rotating shaft (41) of 42. stepping motor (40).

In Fig. 3, 90.-96. The schematic diagram of the imaging of the object (circular spot) at different positions on the optical axis, 97. The optical axis.

In Fig. 5, 21. a row of optical signals, 22. digitized signals corresponding to the optical signals (21), and 23. edge direction data corresponding to the digitized signals (22).

Detailed ways

The invention patent includes two parts: the computer and its camera measurement system shown in Figure 1, and the camera axial displacement device with a high-precision micro-displacement stepping motor as the core shown in Figure 2.

On the computer system (5), run the camera driver program (13) for distribution, and connect the camera (1) to the computer (5) through USB interfaces (4) and (6). Then, let the camera focus and image the object to be measured.

During the measurement process, certain changes in lighting conditions are allowed, as long as the changes do not significantly affect the contrast between light and dark imaged by the measured object. Selecting lighting equipment (15) is helpful to the implementation of the present invention. For example, the uniform lighting method with diffuse reflection, or the intensity of the lighting equipment can be stronger than the influence of ambient stray light. The material of the object to be measured should preferably have more detailed surface reflection characteristics, and this kind of material can also be used as a target to avoid or overcome the smooth reflective surface material.

As shown in Figure 2, the camera (1) is installed on the workbench (33) and moves together with the workbench, and the workbench (33) is socketed with a long threaded rotating shaft (34) with threaded threads. Screw thread rotating shaft (34) is installed on a large workbench (30) by bracing frame (31) and (32), and, thread thread rotating shaft (34) and bracing frame (31) and (32) are all Socketed in the form of a rotating shaft, the threaded rotating shaft (34) can rotate at the socketed place with the support frames (31) and (32) but does not move forward or backward. There is a fixed gear (341) on the threaded rotating shaft (34), which meshes with the gear (42) on the rotating shaft (41) of the stepping motor (40), when the rotating shaft (41) of the stepping motor (40) ) rotates, the rotation of the gear (42) will drive the gear (341) to rotate, and then drive the screw thread rotation shaft (34) to rotate, prompting the camera (1) and its workbench (33) to move forward or backward together . Stepping motor (40) is also installed on the big workbench (30).

Select two output signal lines in the RS232C interface (8) of the computer system (5), FORWARD and BACKWARD as shown in the figure, and a ground wire GROUND are connected to a stepping motor interface circuit (43), here Carry out power amplification, then, be connected to stepper motor (40), control stepper motor (40) to make clockwise rotation or anticlockwise rotation.

The performance of the stepper motor (40) is related to the measurement accuracy of the patent of the present invention, and the motor with fine stepping displacement, high precision and stable operation should be selected for use.

Run the camera shooting and axial displacement measurement program (14) according to the one-dimensional contrast ratio to measure the displacement in real time. The specific steps are described in the "Summary of the Invention", and the main points are described below.

The principle and method of measuring the optical axial micro-displacement of the patent of the present invention is based on whether the focus of optical imaging is clear or not. As shown in Figure 3, the judgment criterion is: when the focus is clear, the observation window area of the image frame has the largest number of image features- - Side reflection condition. The pixel array and its observation window area generated after the photoelectric conversion of the photosensor chip is shown in FIG. 4 . The definition and determination of image features, that is, edge direction data, are shown in Figure 5.

Before the measurement starts, calculate the self-correlation matching coefficient to select a suitable observation window area, judge whether the optical reflection properties of the surface of the measured object are suitable for the measurement work, and optimize the analysis calculation amount.

The coordinate axis selected by the patent of the present invention is not limited to the X axis, and the Y axis is equivalent to it.

The measurement method described in the patent of the present invention is also applicable to other imaging devices.

Claims (3)

1. A method for measuring axial displacement characterized by one-dimensional contrast, which measures axial displacement by a common computer, a camera, camera axial displacement device, a stepper motor and a stepper motor interface circuit, The computer is connected to the camera through its USB interface, and the camera is installed on a camera axial displacement device composed of the stepping motor as the core, and the stepping motor is connected to the RS232C interface of the computer through the stepping motor interface circuit, The camera axial displacement device comprises: the camera (1) is installed on the first workbench (33), the first workbench (33) is socketed with a long screw thread rotating shaft (34) with the screw thread, and the screw The screw thread rotating shaft (34) is installed on a large second workbench (30) by the first support frame (31) and the second support frame (32), and the screw thread rotating shaft (34) and the first support frame (31) and the second support frame (32) are all socketed in the form of a rotating shaft, and the threaded rotating shaft (34) can rotate at the socket with the first support frame (31) and the second support frame (32) But forward or backward displacement does not take place, a fixed gear (341) is arranged on the screw thread rotating shaft (34), and the gear (42) on the rotating shaft (41) of this gear (341) and stepper motor (40) ) interlocking, the stepper motor (40) is also installed on the second workbench (30), and the stepper motor (40) is connected to the RS232C interface (8) of the computer (5) by the stepper motor interface circuit (43); It is characterized in that the method takes pictures with a camera and measures the tiny displacement of the object along the optical axis direction of the camera according to the one-dimensional contrast ratio, including the following steps: Step 1. Take a frame of the image of the measured object in the format of bitmap M×N as the reference frame; take the position of the first pixel in the upper left corner of the pixel array of the frame as the origin, and take the rightward direction as the X-axis direction , the vertical downward direction is the direction of the Y axis, and the unit of the coordinate system is the size of one pixel; a region is selected in the central region of the pixel array with a size of m ₀ ×n ₀ , which is called the observation window, and its distance from The horizontal and vertical edge pixels of the pixel array each have h and v pixels, namely: m ₀ +2h=M, n ₀ +2v=N, where M, N, m ₀ , n ₀ , h, v ∈ positive integer; Step 2. For the pixel array of the above reference frame, derive the side direction data along the X-axis direction and the Y-axis direction pixel by row and pixel by column, and use 3-bit binary values 001, 010 and 100 to represent the positive side, The negative edge and the third type edge constitute two frames of edge direction data {reference _x (x, y)} and {reference _y (x, x, y)}, wherein, the subscript x or y respectively represent the direction of the coordinate axis along, and the symbol "{}" represents the pixels (x, y) in the observation window along the direction of the coordinate axis indicated by the function subscript A collection of edge direction data, save these data; Step 3. For the above two frames of edge direction data, calculate the self-correlation matching coefficient of the pixel array in the observation window in the reference frame: $\mathrm{<span\; class="notranslate">\; auto</span>}<span\; class="notranslate">\; \_</span>{\mathrm{<span\; class="notranslate">\; correlation</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&bull;</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$ $\mathrm{<span\; class="notranslate">\; auto</span>}<span\; class="notranslate">\; \_</span>{\mathrm{<span\; class="notranslate">\; correlation</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&bull;</span>{\mathrm{<span\; class="notranslate">\; reference</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$ In the formula, the operation symbol · represents the binary logic AND operation, and the operation result is either logic 0 or logic 1, and the operation symbol "[]" represents the value corresponding to the value of the logic operation function, or the value 0, or The value is 1, and the combination of parameter variables a and b determines the scale of the associative matching operator array. If a 3×3 associative matching operator array is used: a=-1, 0, 1, b=-1, 0, 1, Therefore, 9 self-correlation matching coefficients auto_correlation _x (a, b) and auto_correlation _y (a, b) will be generated along each coordinate axis direction; Step 4. According to the self-association matching coefficients corresponding to the above two frames of edge direction data, search for the best observation window pixel array that can be matched and compared under the current object surface conditions and lighting conditions: m _x = m ₀ ± step, n _x = n ₀ ± step, 2h = Mm _x , 2v = Nn _x , and m _y = m ₀ ± step, n _y = n ₀ ± step, 2h = Mm _y , 2v = Nn _y , In the formula, the subscripts x and y indicate that their values correspond to the directions along the X-axis and the Y-axis respectively, step is the step parameter in the search process, and the plus and minus signs in front of it are determined by the search direction; take the two sets of values The larger one is the size of the observation window array used for this measurement: m×n; Step 5. For the pixel array of the above-mentioned reference frame, according to its pixel brightness, derive the edge direction data of one frame along the X-axis direction row by row; According to the above-mentioned edge direction data, for the observation window area therein, use accumulators to count the numbers of their corresponding positive and negative sides: N _{X-axis positive} and N _{X-axis negative} , where N represents the number; accumulate the above-mentioned accumulator The counting result and save it, expressed as: N(i, j=0, forw=0, back=0), wherein, i=1, 2, 3..., represents the taken reference frame Sequential counting is also the counting of measurement, j=0, 1, 2, 3,..., represents the counting of the sampling frames taken during the i-th measurement, and the variables forw and back respectively represent the The stepping pulse count corresponding to clockwise rotation and counterclockwise rotation of the stepping motor: forw=0, 1, 2,..., back=0, 1, 2,... , before the start of this measurement, there are: i=1, j=0, forw=0, back=0; Step 6, measurement start: the computer outputs the first forw=1 digital pulse signal to the stepper motor interface circuit through an output control line FORWARD of its RS232C interface, and controls the stepper motor to rotate one step clockwise, and then, Take the sample frame bitmap of j=1 frame; For the pixel array of the above sampling frame, according to its pixel brightness, export its side direction data of one frame along the X-axis direction line by line; According to the above side direction data, for the observation window area, use the accumulator to count the number of their corresponding positive side and negative side: N _{X-axis positive} and N _{X-axis negative} ; accumulate the counting results of the above accumulator and save it , expressed as: N(i, j=1, forw=1, back=0); Step 7, if: N (i, j=1, forw=1, back=0) ≥ N (i, j=0, forw=0, back=0), described computer is through one output of its RS232C interface The control line FORWARD outputs the forwth digital pulse signal to the stepper motor interface circuit: forw=2, 3,..., to control the stepper motor to rotate clockwise and progress, and each clockwise rotation takes a step And analyze and compare the sum N (i, j=forw-1, forw-1, back=0) and N (i, j =forw, forw, back=0), save it, there should be: N(i, j=forw, forw, back=0) ≥ N(i, j=forw-1, forw-1, back =0), until satisfying: N(i, j=forw, forw, back=0)<N(i, j=forw-1, forw-1, back=0), at this moment, record (forw) _MAX = Forw, at the same time, the computer outputs a digital pulse signal to the stepper motor interface circuit through another output control line BACKWARD of its RS232C interface: back=1, controls the stepper motor to rotate one step counterclockwise, at this time, Relative to the initial position before the start of this measurement, the camera is at the best object imaging focus position in this measurement: FocusP=(forw) _MAX -back=(forw) _MAX -1, the calculated result is greater than or equal to 0, indicating that the direction of rotation of the stepper motor is generally clockwise, and the sum of the numbers of all positive and negative sides along the X-axis direction in the observation window of the sampling frame corresponding to the focus position is: N( i, j=(forw) _MAX -1, forw=(forw) _MAX -1, back=0), the total counting result of the sampling frames taken during the measurement is: j=(forw) _MAX +back=(forw) _MAX +1; If: N(i, j=1, forw=1, back=0)<N(i, j=0, forw=0, back=0), said computer is through an output control line BACKWARD of its RS232C interface Output the first back=1 digital pulse signal to the stepper motor interface circuit, control the stepper motor to take a step back in counterclockwise rotation, that is, return to the initial position of this measurement, and then, the computer outputs through its RS232C interface The control line BACKWARD continues to output the back digital pulse signal to the stepper motor interface circuit: back=2, 3,..., to control the stepper motor to further rotate counterclockwise. In the process, each time One step of anti-clockwise rotation is all taken and analyzed and compared before and after each sampling frame observation window along the X-axis direction with the sum of the numbers N(i, j=back+1, forw=1, back) and N( i, j=back, forw=1, back=back-1), there should be: N(i, j=back+1, forw=1, back=back)≥N(i, j=back, forw=1 , back=back-1), save it, and continue like this until it is satisfied: N(i, j=back+1, forw=1, back)<N(i, j=back, forw=1, back-1) , at this moment, record: (back) _MAX =back, simultaneously, described computer outputs a digital pulse signal to described stepping motor interface circuit by another root output control line FORWARD of its RS232C interface: forw=2, controls this The stepper motor rotates one step clockwise. At this time, relative to the initial position before the start of this measurement, the camera is in the best object imaging focus position in this measurement: FocusP=forw-(back) _MAX =2-( back) _MAX , the calculation result is a negative value, indicating that the rotation direction of the stepper motor is anti-clockwise, and the number of positive and negative sides along the X-axis direction in the sampling frame observation window corresponding to the focus position The sum is: N(i, j=((back) _MAX -1)+1=(back) _MAX , forw=1, back=(back) _MAX -1), the total count of sampled frames taken during the measurement The result is: j=forw+(back) _MAX =2+(back) _MAX ; To sum up, relative to the initial position before the start of this measurement, the displacement of the object obtained in this measurement along the optical axis of the camera is: Δz(i)=forw-back, The values of the counters forw and back in the above calculation formula are the final counting results in this measurement process, and the calculation results have positive or negative signs, which respectively indicate the direction of the displacement that occurs along the optical axis to advance or retreat, corresponding to the The direction of rotation of the stepper motor is clockwise or counterclockwise, which is determined by the arrangement of the camera axial displacement device composed of the stepper motor as the core; The total displacement is: ΔZ ₀ (i)=ΔZ ₀ (i-1)+Δz(i), Among them, ΔZ ₀ (i-1) is the accumulated axial displacement before this measurement; Step 8. Prepare for the next measurement: add 1 to the counter i of the number of measurements, and take the values of all positive and negative sides along the X-axis direction in the corresponding object imaging frame observation window at the best object imaging focus position determined by the ith measurement The sum of the numbers is used as the new measurement reference value: N(i, j=0, forw=0, back=0)=N(i, j=(forw) _MAX -1, forw=(forw) _MAX -1, back=0), or N(i, j =0, forw=0, back=0)= N(i, j=((back) _MAX -1)+1=(back) _MAX , forw=1, back=(back) _MAX -1); Step 9, jump to step 6, continue to measure. 2. the method for measuring axial displacement characterized by one-dimensional contrast according to claim 1, characterized in that, the definition of side direction data in said step 2, step 5 and step 6 is: In the pixel array, along the X-axis or along the Y-axis, if the light intensity value of a pixel is smaller than the corresponding light intensity value of the second pixel behind it by an error tolerance value error, that is, if: I(X,Y)<I(X+2,Y)-error or I(X,Y)<I(X,Y+2)-error, It is defined that there is a positive edge along the axis between these two pixels; if the light intensity value of a pixel is greater than the corresponding light intensity value of the second pixel behind it by an error tolerance value error, that is, if : I(X, Y)>I(X+2, Y)+error or I(X, Y)>I(X, Y+2)+error, Then it is defined that there is a negative edge along the axis between these two pixels; the edge obtained in this way is located at the position of the first pixel after this pixel, that is, the pixel located in the middle of the two pixels involved in the comparison ; If a certain light intensity value of a pixel is close to the corresponding light intensity value of the second pixel behind it, the difference between the values does not exceed an error tolerance value error, that is, if: I(X+2,Y)-error≤I(X,Y)≤I(X+2,Y)+error, or I(X, Y+2)-error≤I(X,Y)≤I(X,Y+2)+error, Then it is considered that there is no corresponding "edge" along the axis between the two pixels, or it is called the third type of edge; in the above expression, the independent variable X and the independent variable Y represent the pixel in question in the pixel array The coordinates in the X-axis and Y-axis directions; Along a certain coordinate axis direction, all positive edges, negative edges, and third-type edges of the corresponding pixel row or pixel column form the edge direction data of the row or column along the direction of the coordinate axis; according to the specific lighting conditions, pre-set Set the error tolerance value error in the above formula to a small value; there is no edge direction data at the pixel positions on the four sides and corners in the pixel array. 3. the method for measuring axial displacement characterized by one-dimensional contrast according to claim 1, characterized in that, the method of searching for the best observation window pixel array in the step 4 comprises: For the observation window of the pixel array and the k × k associated matching operator array (a, b), k∈ positive integer, k × k self-associated matching coefficients will be generated along a certain coordinate axis direction, and these are compared according to the following inequality Self-correlation matching coefficient: auto_correlation(a,b)≥auto_correlation(0,0)×similarity In the formula, similarity describes the similarity between the observation window and its neighboring pixel arrays of the same size. The similarity is pre-set and debugged and selected according to the lighting conditions and the texture of the surface of the measured object; If there are more than k×k×1/3 self-correlation matching coefficients satisfying the above inequality, it is necessary to expand the scope of the observation window for each step row and step column: let m=m ₀ +step, n=n ₀ +step, and recalculate Newly observe the self-correlation matching coefficients of the window, and carry out the above comparison until the self-correlation matching coefficients satisfying the above inequality are no more than k×k×1/3, at this time, 2h=Mm, 2v=Nn, where, step It is a stepping parameter, the initial value is 1, and it increases by 1 every time the scale of the observation window needs to be expanded; if it exceeds a predetermined range in the frame and no suitable observation window has been found, it is considered that the texture of the reflective surface of the object is not suitable For measurement work, and give prompts and warnings; If there are no more than k×k×1/3 self-association matching coefficients that satisfy the above inequality, it means that the surface structure features of the object to be photographed are fine enough, and the values between the nearest adjacent pixels can be distinguished, and further try to narrow the scope of the observation window Each step row and step column to reduce the calculation workload: set m=m ₀ -step, n=n ₀ -step, recalculate the self-correlation matching coefficient of the observation window, and perform the above comparison, and the step parameter step increases each time 1. Until the number of self-correlation matching coefficients satisfying the above inequality in the selected observation window area is not less than k×k×1/3, at this time, it is considered that the optimal observation window pixel array has been searched.

Images