TWI397668B - System and method for correcting an image - Google Patents

Description

Image correction system and method

The invention relates to an image correction system and method.

In precision image measurement, the standard industrial charge coupled device (CCD) with a special lens can get high-definition images, but due to CCD and lens characteristics, the captured image and the real image are in high error amplification ratio. There will be a certain deformation in the lower relative ratio, which has a certain influence on the precision measurement results. Referring to FIG. 1 and FIG. 2, the actual image of the one-line grid pattern and the image imaged by the standard CCD are respectively visible, and the line grid in FIG. 2 is somewhat deformed. In order to overcome such mechanical image distortion, it is necessary to provide a method for correcting images.

In view of the above, it is necessary to propose an image correction system and method which uses a grid correction method to compensate for mechanical errors and correct images.

An image correction system includes a computer and an image measuring machine. The computer includes an image capture card, and the image measuring machine is equipped with a charge coupled device. The charge coupled device is coupled with an industrial optical lens for acquiring an image of the workpiece to be tested, and transmitting the image to the image. The image of the computer is captured in the card. The computer further includes: an image correction program for correcting the image of the workpiece to be tested, the image correction program comprising: a deviation value calculation module for calculating a deviation value of the measurement reference point; and an image correction module For the bias of the measurement reference point calculated according to the above The difference is calculated on the point on the image of the workpiece to be measured.

Further, the measurement reference point is a grid intersection of a plurality of grid patterns into which the charge coupling device is divided into an imaging range of the industrial optical lens; the deviation value of the measurement reference point is the measurement The difference between the actual value of the reference point and the measured value.

An image correction method for correcting an image of a workpiece to be tested measured by a charge coupled device of an image measuring machine by a computer. The method includes the steps of: selecting a point P0 on the image; converting the mechanical coordinate of the point P0 to a coordinate of the charge coupled device; acquiring four measurement reference points closest to the point P0; calculating the point P0 to The distances D0, D1, D2, D3 of the above four measurement reference points; the deviation values A0, A1, A2, A3 of the above four measurement reference points are obtained; D0, D1, D2, D3 and the obtained deviations are calculated according to the above The values A0, A1, A2, A3 calculate the deviation value of the point P0; and perform correction calculation on the point P0 based on the deviation value.

Further, the measurement reference point is a grid intersection of a plurality of grid patterns into which the charge coupling device on the image measuring machine is divided into the imaging range of the industrial optical lens.

Compared with the prior art, the image correction system and method provided by the present invention adopts a grid correction method to compensate and correct the image, thereby effectively eliminating or reducing the influence of the error of the machine itself.

Referring to Figure 3, there is shown a hardware architecture diagram of a preferred embodiment of the image correction system of the present invention. The image correction system includes a computer 1 and a workpiece 5 to be tested. Image measuring machine 6. The Z-axis of the image measuring machine 6 is further equipped with a Charged Coupled Device (CCD) 7 for collecting continuous images, and the CCD 7 is equipped with an industrial optical lens. The CCD 7 can be used to image the workpiece 5 to be tested in conjunction with the industrial optical lens. Further, for the purpose of image correction, a calibration piece 4 is placed under the workpiece 5 to be tested, which is equivalent to a scaled and highly accurate ruler, which can be used to measure the workpiece to be tested. The actual length of 5. The computer 1 is equipped with an image capture card 10 and an image correction program 11. The CCD 7 is connected to the image capturing card 10 through the image data line, and the image of the workpiece 5 to be tested acquired from the image measuring machine 6 is transmitted to the image capturing card 10 and displayed on the display screen of the computer 1 ( Not shown). The image correction program 11 is mainly used for performing correction calculation on the acquired image of the workpiece 5 to be tested.

Referring to FIG. 4, it is a functional module diagram of the image correction program 11 in FIG. Each module referred to in the present invention is a program segment of the image correction program 11 that performs a specific function, and is more suitable for describing the execution process of the software in the computer than the program itself. Therefore, the description of the software in the present invention is a module. description.

The image correction program 11 mainly includes an offset value calculation module 110 and an image correction module 111.

The deviation value calculation module 110 is configured to calculate a deviation value of the measurement reference point. In a preferred embodiment of the present invention, the grid error is used to correct the mechanical error and correct the image. The grid line complements the CCD 7 and the industrial optical lens respectively under different objective lenses and different magnifications. In general, the CCD 7 can be adjusted to 1× objective lens and 2× objective lens. In this case, the industrial optical lens can be used in the case of 1x magnification, 2x magnification and 3x magnification, so the CCD7 can be used in combination with the industrial optical lens in six combinations: 1x objective lens with 1x magnification, 1x objective lens with 2x magnification, 1x objective with 3x magnification, 2x objective with 1x magnification, 2x objective with 2x magnification and 2x objective with 3x magnification. In the case of each combination, the range in which it can be imaged is divided into squares as shown in Fig. 2, where the intersection of the grid lines is the measurement reference point for image correction. The deviation value of the measurement reference point is the difference between the actual value of the measurement reference point (ie, the actual coordinate value) and the measured value (ie, the measured coordinate value). The deviation value of the measurement reference point in each group combination can be recorded in a position corresponding to the table shown in FIG. 7.

The image correction module 111 is configured to perform correction calculation on a point on the image of the workpiece 5 to be measured according to the calculated deviation value of the measurement reference point. When the image is corrected, if the point on the image falls on the intersection of the grid line, that is, the measurement reference point, the point can be directly corrected by using the corresponding deviation value recorded in the table. That is, the actual value of the point is calculated by using the measured value of the point plus the deviation value of the record corresponding to the point in the table; if the point on the image falls inside the square, the distance is transmitted from the point The nearest four grid intersections calculate the offset of the point, and correct the point. The correction method is as follows: first find the four grid intersections closest to the point; then calculate the point to the above four The distances D0, D1, D2, D3 of the intersections of the grid lines; the deviation values A0, A1, A2, A3 of the intersections of the four grid lines are obtained from the above table;

Calculate the deviation value A of the point; finally, use the deviation value A of the point plus the measured value of the point to calculate the actual value of the point, and complete the correction of the point.

Referring to FIG. 5, it is a flowchart of an implementation of calculating the deviation value of the measurement reference point by using the deviation value calculation module 110 in the preferred embodiment of the image correction method of the present invention.

Usually, after an image measuring machine 6 is produced and the components on it (such as CCD7, etc.) are fixed, the relevant calibration engineer will calculate the CCD7 and the industrial optical lens respectively in different objective lenses (1x objective lens and 2x objective lens) The deviation value of the measurement reference point in the case of a combination of different magnifications (1x magnification, 2x magnification, and 3x magnification), and the deviation value of the measurement reference point in each group of matching cases is recorded in In the table shown in Figure 7. The preferred embodiment is described by taking the case of a CCD 7 with a 1x objective lens and a 1x magnification industrial optical lens as an example. The implementation steps are as follows:

In step S100, the CCD 7 is adjusted to a 1× objective lens and the industrial optical lens is adjusted to 1× magnification, and the imaging range thereof is divided into several square patterns as shown in FIG. 2 .

In step S101, an initial table is generated according to the above-mentioned checkered pattern for recording the deviation value of each grid intersection. The initial value of each deviation value in the table is 0.

In step S102, a grid intersection is selected as the measurement reference point.

In step S103, the coordinates of the point are converted into coordinates of the CCD 7. The conversion includes: conversion of coordinate systems and conversion of units. The coordinate system is converted by converting the mechanical coordinate system indicated by the upper left corner of the table 6 as shown in FIG. 3 to the center of the CCD 7, and the X-axis direction is the same as the mechanical coordinate system, and the Y-axis direction. The coordinate system opposite to the above-mentioned mechanical coordinate system; the unit conversion is to convert the dimensional unit of the mechanical coordinate system into a unit unit of the CCD 7 coordinate system by a fixed ratio value.

Step S104, the image correction program 11 calculates the measured value of the point under the coordinate system of the CCD 7 (ie, the measured coordinate value), and the calibration piece 4 measures the actual value of the point under the coordinate system of the CCD 7 (ie, actual Coordinate value).

Step S105, calculating the measurement reference point deviation value, that is, calculating the difference between the actual value of the measurement reference point and the measured value.

In step S106, the deviation value of the point is recorded in the position corresponding to the initial table.

In step S107, it is determined whether all the intersections of the ruled lines have been calculated.

If not all the points have been measured, the process returns to step S102, and the next ruled line intersection is selected as the measurement reference point.

If all the points have been measured, the process ends.

After the above process ends, the table for recording the deviation value of each grid line intersection point can be saved in a special encryption file, and the relevant correction engineer needs to input the decryption password to open the table and view or modify the form data. To prevent other users from accidentally accessing the information in the form. Modifications increase the calibration workload of the calibration engineer.

Similarly, the CCD7 with 1x objective lens is matched with 2x magnification industrial optical lens, 1x objective lens CCD7 with 3x magnification industrial optical lens, 2x objective lens CCD7 with 1x magnification industrial optical lens, 2x objective lens In the case where the CCD 7 is equipped with a 2x magnification industrial optical lens and a 2x objective lens CCD7 with a 3x magnification industrial optical lens, the flow for calculating the deviation value of the measurement reference point is the same as the flow described above. Therefore, by repeatedly performing the above process, six tables can be generated corresponding to each combination of the CCD 7 and the industrial optical lens.

Generally, in the actual measurement work, the image measuring machine 6 may move from one laboratory to another according to the needs of the measurement or may take a long time from the last calibration, which may cause the machine to be caused. The upper component deviates from the original position, which in turn causes the calculated deviation value in the above table to be inaccurate, so the deviation value recorded in the above table is corrected. The flow of the offset value correction is shown in Figure 6.

FIG. 6 is a flow chart showing an implementation of correcting the deviation value of the measurement reference point by using the deviation value calculation module 110 in the preferred embodiment of the image correction method of the present invention. The present invention is described by taking a measurement reference point corresponding to the case where the CCD 7 of the 1x objective lens is matched with the industrial optical lens of 1× magnification as an example. First, in step S200, the correction engineer inputs the decryption code. Generally, in order to better protect the form data, multiple passwords can be set for the encrypted file.

In step S201, it is verified whether the password is valid. If the password is invalid, in step S202, the password is incorrect, and the process returns to step S200.

If the password is valid, the encrypted file is untied, and in step S203, the correction engineer selects a form to be corrected. In the present embodiment, the correction engineer selects a table corresponding to the 1x objective lens CCD7 and the 1x magnification industrial optical lens.

In step S204, a measurement reference point is selected, that is, a grid line intersection is corrected. The correction of the intersection of the grid lines is performed by recalculating the deviation value of the intersection of the grid lines. The method is the same as the method for calculating the deviation value in FIG. 5, comprising: converting the mechanical coordinate value of the selected grid intersection to the coordinate value of the CCD 7, measuring the actual value of the reference point, and calculating the measured value of the point; Calculate the deviation value for this point.

Step S205, modifying the deviation value of the measurement reference point at a position corresponding to the selected table.

Step S206, saving the modified table.

According to the needs of the calibration, after the calibration of the point is completed, the calibration engineer can select other measurement reference points to perform correction calculation according to this procedure. At the same time, after the data of the form is completely corrected, the other five tables can also be selected to be corrected according to this process.

Referring to Figure 7, there is shown a schematic diagram of the form data in the preferred embodiment of the present invention. The table records the deviation values of the reference points on the X-axis and the Y-axis separately in two tables. A horizontal line of data represented by the letter Y as shown in the figure is used as a dividing line. In the above part of the table, a horizontal line of the letter X indicates the actual value of each reference point in the X-axis direction, and a vertical column headed by the letter X. The number indicates the actual value in the Y-axis direction corresponding to each reference point. The intersection of the actual value of a point in the X-axis direction and the actual value in the Y-axis direction is the point on the X-axis. The deviation value of the direction. Similarly, in the following part of the table, a horizontal line of the letter Y indicates the actual value of each of the reference points in the X-axis direction, and a vertical column number represented by the letter Y indicates the corresponding reference point in the Y-axis direction. The actual value, the intersection of the actual value of a point in the X-axis direction and the actual value of the Y-axis direction is the deviation value of the point in the Y-axis direction.

For example, the actual value of a reference point as shown in FIG. 7 is (0, 0). As can be seen from the table data, the deviation values of the point in the X-axis direction and the Y-axis direction are -0.0007 and 0.0002, respectively.

Referring to FIG. 8 , it is a flowchart of an implementation of correcting and calculating an image by using the image correcting module 111 in the preferred embodiment of the image correcting method of the present invention. The present invention is described by taking an example of correcting a point on an image. First, in step S300, a point P0 on the image of the workpiece 5 to be tested is selected for calibration measurement.

In step S301, the mechanical coordinate of the point P0 is converted into a CCD coordinate. The conversion is the same as S103 in FIG. 5, including: converting the mechanical coordinate system of the machine table 6 to the center of the CCD 7 as the origin, the X-axis direction is the same as the mechanical coordinate system, and the Y-axis direction and the mechanical coordinate system are The opposite coordinate system; and converts the dimensional unit of the mechanical coordinate system into a primitive unit of the CCD7 coordinate system at a fixed ratio value.

In step S302, four grid intersections closest to the point P0 are calculated. Referring to FIG. 9, it is a schematic diagram of calculating four grid line intersections closest to the point P0. Wherein, the CCD coordinates of the point P0 are (0.5, 0.5), and the image correction module 111 calculates the intersection of the grid lines of the four squares closest to the point P0, and it can be calculated that the point P0 falls within the square ABCD.

In step S303, the image correction module 111 calculates the distances D0, D1, D2, and D3 of the four vertices from the point P0 to the square ABCD.

In step S304, the deviation value A of the point P0 is calculated.

The formula for calculating the deviation value of point P0 is:

Among them, A0, A1, A2, and A3 respectively represent the deviation values of the four vertices of the square ABCD recorded in the above table. When calculating the deviation value of the point P0 in the X direction, A0, A1, A2, and A3 are squares, respectively. The deviation value of the four vertices of the ABCD in the X-axis direction. Similarly, when calculating the deviation value of the point P0 in the Y direction, A0, A1, A2, and A3 are the deviation values of the four vertices of the square ABCD in the Y-axis direction, respectively.

To convert the above formula, you can generally multiply the formula by D0*D1*D2*D3 and convert the formula to: If you remember:

Then the formula can be reduced to:

Therefore, the deviation value A of the point P0 is calculated.

In step S305, the correction calculation is performed on the point P0 according to the calculated deviation value A. The correction is obtained by using the measured value of the point P0, that is, the above-mentioned CCD coordinate value (0.5, 0.5), and adding the deviation value A of the point to obtain the actual value of the point.

The preferred embodiment of the present invention describes a relatively complete process for calculating the deviation value of the measurement reference point and correcting the point on the image according to the deviation value of the measurement reference point, when actually correcting the image of a workpiece to be tested If there is already a table as shown in Figure 7 that records the deviation value of each measurement reference point, and the deviation value of each measurement reference point is accurate, you can directly use the data in the table to implement as shown in Figure 8. The process corrects the image without the deviation value calculation module 110 recalculating the deviation value of each measurement reference point.

The image correction system and method provided by the invention adopts the method of grid correction to compensate and correct the image, effectively eliminating or reducing the influence of the error of the machine itself; and storing the correction data in a special encryption file can prevent The user can modify it at will.

The above is only the preferred embodiment of the present invention, and has been used in a wide range of applications. Any other equivalent changes or modifications which are not departing from the spirit of the present invention should be included in the following claims. Inside.

1‧‧‧ computer

4‧‧‧ calibration film

5‧‧‧Workpiece

6‧‧‧Image measuring machine

7‧‧‧Charge-coupled device

10‧‧‧Image Capture Card

11‧‧‧Image Correction Program

110‧‧‧ Deviation value calculation module

111‧‧‧Image Correction Module

S300‧‧‧Select a point on the image P0

S301‧‧‧ Convert point P0 to CCD coordinates

S302‧‧‧ Calculate the four grid intersections closest to point P0

S303‧‧‧ Calculate the distance from point P0 to the intersection of the above four grid lines D0, D1, D2, D3

S304‧‧‧ Calculate the deviation value A of point P0

S305‧‧‧Correct calculation for point P0

Figure 1 is the actual image of a one-frame graphic.

FIG. 2 is an image of the square pattern in FIG. 1 imaged by a CCD.

3 is a hardware architecture diagram of a preferred embodiment of the image correction system of the present invention.

FIG. 4 is a functional block diagram of the image correction program in FIG. 3.

FIG. 5 is a flow chart showing an implementation of calculating a deviation value of a measurement reference point in a preferred embodiment of the image correction method of the present invention.

6 is a flow chart showing an implementation of correcting a deviation value of a measurement reference point in a preferred embodiment of the image correction method of the present invention.

Figure 7 is a schematic illustration of a table in a preferred embodiment of the invention.

FIG. 8 is a flow chart showing an implementation of correcting an image in a preferred embodiment of the image correcting method of the present invention.

Figure 9 is a schematic diagram of calculating the intersection of the nearest grid lines at a point on the distance image.

S300‧‧‧Select a point on the image P0

S301‧‧‧ Convert point P0 to CCD coordinates

S302‧‧‧ Calculate the four grid intersections closest to point P0

S303‧‧‧ Calculate the distance from point P0 to the intersection of the above four grid lines D0, D1, D2, D3

S304‧‧‧ Calculate the deviation value A of point P0

S305‧‧‧Correct calculation for point P0

Claims (6)

An image correction system includes a computer and an image measuring machine, the computer includes an image capturing card, the image measuring machine is mounted with a charge coupled device, and the charge coupled device is coupled with an industrial optical lens for acquiring The image of the workpiece to be tested is transmitted to the image capture card of the computer, wherein the computer further includes: an image correction program for correcting the image of the workpiece to be tested, the image correction program includes: a deviation value calculation module, configured to calculate a deviation value of the measurement reference point, wherein the measurement reference point is a grid line intersection of the plurality of grid patterns into which the imaging range of the charge coupled device is matched with the industrial optical lens a point, and a deviation value of the measurement reference point is a difference between an actual value and a measured value of the measurement reference point; and an image correction module configured to perform an image on the image of the workpiece to be measured according to the deviation value of the measurement reference point calculated above Point to perform correction calculation. The image correction system of claim 1, wherein the deviation value of the measurement reference point is stored in a table. The image correction system of claim 1, wherein the system further comprises a calibration sheet for measuring an actual value of the measurement reference point. An image correction method for correcting an image of a workpiece to be tested measured by a charge coupled device of an image measuring machine by using a computer, the method comprising the steps of: selecting a point P0 on the image; a mechanical coordinate converted to a coordinate of the charge coupled device; acquiring four measurement reference points closest to the point P0, wherein the measurement The reference point is a grid intersection of a plurality of grid patterns into which the charge coupling device on the image measuring machine is divided into an imaging range of the industrial optical lens; and the distance D0 from the point P0 to the above four measurement reference points is calculated, D1, D2, D3; obtaining the deviation values A0, A1, A2, A3 of the above four measurement reference points, wherein the deviation value of the measurement reference point is the difference between the actual value and the measured value of the measurement reference point; The calculated D0, D1, D2, D3 and the acquired deviation values A0, A1, A2, A3 calculate the deviation value of the point P0; and perform correction calculation on the point P0 according to the deviation value. The image correction method of claim 4, wherein the obtaining the deviation value of the measurement reference point comprises: selecting a grid intersection as the measurement reference point for measuring; converting the mechanical coordinate of the measurement reference point a coordinate of the charge coupled device; calculating a measured value of the measurement reference point; measuring an actual value of the measurement reference point by using a calibration slice; and calculating a difference between the actual value and the measured value to obtain the measurement reference point Deviation. The image correction method of claim 5, wherein the deviation value of the measurement reference point is stored in a table.