Vision measurement system based on bilinear interpolation error compensation

Abstract

The invention discloses a vision measurement system based on bilinear interpolation error compensation, and relates to the field of vision measurement. Comprising a visual acquisition module, a man-machine interaction module, a motion platform and a data processing module, and the following steps are realized through the above modules: S1, carrying out image acquisition, then obtaining user input through the man-machine interaction module, moving the motion platform to a designated position, and recording a pose in real time; s2, the data processing module processes, corrects and compensates the acquired image; and S3, outputting a high-precision measurement result. The method has the advantages that a bilinear interpolation method is adopted, nonlinear errors are effectively processed, the camera position and angle change can be more accurately adapted, and therefore the measurement precision is improved, error compensation is automatically processed through an algorithm, manual intervention is reduced, operation is easier and more convenient, the usability and reliability of the system are improved, and the method is suitable for large-scale popularization and application. And errors caused by human factors are reduced.

Description

Technical Field

[0001] The invention relates to the field of visual measurement, and in particular to a visual measurement system based on bilinear interpolation error compensation. Background Art

[0002] In the field of visual measurement, existing technologies mainly rely on geometric models based on camera calibration and motion platforms for measurement. These technologies usually use traditional calibration methods, such as Zhang Zhengyou calibration method, to obtain the internal and external parameters of the camera. These methods can provide relatively accurate measurement results in some cases, but in practical applications, they still face the following problems:

[0003] 1. Error accumulation: Due to the errors in the camera, motion platform and algorithm calibration, the errors gradually accumulate during the measurement process, especially in long-term or high-frequency measurements;

[0004] 2. High complexity: Existing methods require multiple manual adjustments and calculations during the calibration process, which is complex and error-prone;

[0005] 3. Lack of flexibility: Existing technologies often have difficulty adapting to changes in dynamic environments and cannot quickly respond to errors under different measurement conditions;

[0006] The existence of the above problems is mainly due to the physical characteristics of the camera and the motion platform and their interaction with the environment. It is difficult to completely eliminate the errors, especially in high dynamic range or complex background. The recognition and matching of feature points will be affected, resulting in a decrease in accuracy. Traditional algorithms usually assume that the errors are linear, while the actual errors are often nonlinear. Therefore, it is difficult to achieve the expected effect when using linear models for compensation. Based on this, the present invention designs a visual measurement system based on bilinear interpolation error compensation. Summary of the invention

[0007] The purpose of the present invention is to solve the problems in the prior art and to propose a visual measurement system based on bilinear interpolation error compensation.

[0008] A visual measurement system based on bilinear interpolation error compensation includes the following modules:

[0009] A visual acquisition module, wherein the visual acquisition module includes an image acquisition module, a light source and a light source controller, wherein the light source controller adjusts the light source to achieve a suitable light intensity to ensure that the image acquisition module can acquire high-quality images under different light conditions;

[0010] A human-computer interaction module, the human-computer interaction module comprising a display and an industrial computer;

[0011] The motion platform includes a communication module and a control system, which is responsible for controlling the movement of the calibration plate and recording the posture data of each position in real time. The control system ensures the accuracy of the motion, and the communication module realizes data exchange with other modules;

[0012] The data processing module includes a software module and a data processing module. The data processing module includes:

[0013] Calibration algorithm, used to obtain the internal and external parameters of the camera to ensure the accuracy and stability of the measurement system in a dynamic environment;

[0014] Image processing algorithm, extracts feature points in the image, generates data for error calculation, and ensures that the acquired feature points have high reliability;

[0015] Distortion correction algorithm, which performs distortion correction on the acquired images to improve image quality and ensure the reliability of subsequent processing;

[0016] Reprojection algorithm calculates the reprojection position of feature points in the world coordinate system, provides a basis for error calculation, and ensures the accuracy of measurement results;

[0017] Bilinear compensation algorithm, based on the error of known points, compensates through bilinear interpolation to further improve measurement accuracy;

[0018] At the same time, the following steps are implemented through the above modules:

[0019] S1: Perform image acquisition, then obtain user input through the human-computer interaction module, move the motion platform to the specified position, and record the position and posture in real time;

[0020] S2: The data processing module processes, corrects and compensates the acquired images;

[0021] S3: Output high-precision measurement results.

[0022] In the above-mentioned visual measurement system based on bilinear interpolation error compensation, the distortion correction algorithm obtains the internal and external parameters of the camera by extracting and calibrating image feature points, and extracts and calibrates feature points based on the image obtained by image acquisition to obtain the internal and external parameters of the camera.

[0023] In the above-mentioned visual measurement system based on bilinear interpolation error compensation, the calibration coordinate system solves the matrix of the machine and the camera, puts the calibration plate into the motion platform, performs translation operation, records 10-15 sets of motion coordinates RealMacPos of the machine and the corresponding calibration plate coordinates PixPos after distortion correction, and obtains the homography matrix through affine transformation.

[0024] In the above-mentioned vision measurement system based on bilinear interpolation error compensation, the purpose of the reprojection algorithm is to convert the image coordinate system to the interpolation of the coordinates CalMacPos and RealMacPos of the machine coordinate system as the reference error, where:

[0025] CalMacPos=M·PixPos.

[0026] In the above-mentioned visual measurement system based on bilinear interpolation error compensation, when the bilinear interpolation compensation algorithm is used, the error values ​​in the x and y directions are used as reference values ​​for reverse compensation so that the error tends to 0. The specific steps are as follows:

[0027] Step 1: Select four nearby known measurement points and their corresponding reprojection errors v 00 ,v 01 ,v 10 ,v 11 ;

[0028] Step 2: Use the bilinear interpolation formula to calculate the compensation value of the target position based on the error value of the known point. The model is as follows:

[0029]

[0030] Where v(x,y) represents the error value of the (x,y) position, X and Y are the lengths of the sides in the x and y directions respectively;

[0031] The measurement obtains the current image position PixPos1 by image processing and calculates:

[0032] MacPos=M·PixPos1

[0033] Step 3: Update the position by the compensation value to obtain the new measurement coordinates:

[0034] NewMacPos=MacPos+v.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] 1. Nonlinear error processing: The present invention adopts a bilinear interpolation method to effectively process nonlinear errors and can more accurately adapt to changes in camera position and angle, thereby improving measurement accuracy.

[0037] 2. Dynamic response capability: The system can calculate compensation values ​​in real time, significantly reducing error accumulation, especially in dynamic environments, and can quickly respond to changes in different measurement conditions to maintain measurement stability.

[0038] 3. Simplify the operation process: Through the algorithm to automatically handle error compensation, it reduces human intervention, makes the operation easier, improves the usability and reliability of the system, and reduces errors caused by human factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a framework diagram of the visual measurement system design in a visual measurement system based on bilinear interpolation error compensation proposed by the present invention.

[0040] Figure 2 This is a flow chart of a visual measurement system in a visual measurement system based on bilinear interpolation error compensation proposed by the present invention.

[0041] Figure 3 This is an error analysis diagram of a visual measurement system based on bilinear interpolation error compensation proposed by the present invention.

[0042] Figure 4 This is a schematic diagram of the bilinear interpolation principle in a visual measurement system based on bilinear interpolation error compensation proposed by the present invention. DETAILED DESCRIPTION

[0043] Reference Figure 1-4 , a visual measurement system based on bilinear interpolation error compensation, comprising the following steps:

[0044] S1: Perform image acquisition, then obtain user input through the human-computer interaction module, and the motion platform moves to

[0045] Specify the location and record the posture in real time;

[0046] S2: The data processing module processes, corrects and compensates the acquired images;

[0047] S3: Output high-precision measurement results.

[0048] The above steps include the following modules:

[0049] The visual acquisition module includes an image acquisition module, a light source and a light source controller. The light source controller adjusts the light source to achieve a suitable light intensity to ensure that the image acquisition module can obtain high-quality images under different lighting conditions.

[0050] Human-computer interaction module, which includes a display and an industrial computer;

[0051] Motion platform, which includes a communication module and a control system. It is responsible for controlling the movement of the calibration plate and recording the posture data of each position in real time. The control system ensures the accuracy of the movement, and the communication module realizes data exchange with other modules.

[0052] Data processing module, the data processing module includes a software module and a data processing module, the data processing module includes:

[0053] The calibration algorithm is used to obtain the internal and external parameters of the camera to ensure the accuracy and stability of the measurement system in a dynamic environment. The calibration algorithm solves the matrix between the machine and the camera. The specific steps include placing the calibration plate on the motion platform, performing translation operations, recording 10-15 sets of motion coordinates RealMacPos of the machine and the corresponding calibration plate coordinates PixPos after distortion correction, and obtaining the homography matrix through affine transformation;

[0054] Image processing algorithm, extracts feature points in the image, generates data for error calculation, and ensures that the acquired feature points have high reliability;

[0055] The distortion correction algorithm performs distortion correction on the acquired images to improve the image quality and ensure the reliability of subsequent processing. The distortion correction algorithm obtains the internal and external parameters of the camera by extracting and calibrating the image feature points, and extracts and calibrates the feature points based on the images obtained by image acquisition to obtain the internal and external parameters of the camera.

[0056] The image acquisition consists of a CCD camera, a light source, and a light source controller. Adjust the appropriate light intensity and place the calibration plate on the motion platform to ensure that the field of view occupied by the calibration plate is greater than 1 / 3 of the camera field of view. The control strategy of the motion platform is: move between multiple predetermined positions, collect data according to the set path, and ensure that each position data corresponds to the camera image one by one, and take 12 sets of images, as shown in Table 1;

[0057] Table 1: 12 location data information tables

[0058]

[0059] The calibration algorithm is used to obtain the internal and external parameters of the camera to ensure the accuracy and stability of the measurement system in a dynamic environment. The calibration coordinate system is to solve the matrix between the machine and the camera. The calibration plate is placed on the motion platform and translated. The 10-15 sets of motion coordinates RealMacPos of the machine and the corresponding calibration plate coordinates PixPos after distortion correction are recorded. The homography matrix M is obtained through affine transformation as shown in Table 2.

[0060] Table 2: Coordinate system transformation matrix M

[0061] -0.00113001 -0.104213 -187.559 -0.104318 0.00141322 597.448 0 0 1

[0062] The reprojection algorithm calculates the reprojection position of the feature point in the world coordinate system, provides a basis for error calculation, and ensures the accuracy of the measurement results. The purpose of the reprojection algorithm is to convert the image coordinate system to the interpolation of the coordinates CalMacPos and RealMacPos of the machine coordinate system as the reference error, where:

[0063] CalMacPos=M·PixPos

[0064] The bilinear compensation algorithm uses bilinear interpolation to compensate for the errors of known points to further improve the measurement accuracy. When using the bilinear interpolation compensation algorithm, the error values ​​in the x and y directions are used as reference values ​​for reverse compensation to make the error approach 0. The specific steps are as follows:

[0065] Step 1: Select four nearby known measurement points and their corresponding reprojection errors v 00 ,v 01 ,v 10 ,v 11 ;

[0066] Step 2: Use the bilinear interpolation formula to calculate the compensation value of the target position based on the error value of the known point. The model is as follows:

[0067]

[0068] Where v(x,y) represents the error value of the (x,y) position, and X and Y are the side lengths in the x and y directions, respectively.

[0069] The current image position PixPos1 is obtained by image processing and calculated:

[0070] MacPos=M·PixPos1

[0071] Step 3: Update the position by the compensation value to obtain the new measurement coordinates:

[0072] NewMacPos=MacPos+v.

[0073] It is known from common technical knowledge that the present invention can be implemented by other embodiments that do not deviate from its spirit or essential features. Therefore, the above disclosed embodiments are only illustrative in all respects and are not exclusive. All changes within the scope of the present invention or within the scope equivalent to the present invention are included in the present invention.

Claims

1. A visual measurement system based on bilinear interpolation error compensation, characterized in that: Includes the following modules: A visual acquisition module, wherein the visual acquisition module includes an image acquisition module, a light source and a light source controller, wherein the light source controller adjusts the light source to achieve a suitable light intensity to ensure that the image acquisition module can acquire high-quality images under different light conditions; A human-computer interaction module, the human-computer interaction module comprising a display and an industrial computer; The motion platform includes a communication module and a control system, which is responsible for controlling the movement of the calibration plate and recording the posture data of each position in real time. The control system ensures the accuracy of the motion, and the communication module realizes data exchange with other modules; The data processing module includes a software module and a data processing module. The data processing module includes: Calibration algorithm, used to obtain the internal and external parameters of the camera to ensure the accuracy and stability of the measurement system in a dynamic environment; Image processing algorithm, extracts feature points in the image, generates data for error calculation, and ensures that the acquired feature points have high reliability; Distortion correction algorithm, which performs distortion correction on the acquired images to improve image quality and ensure the reliability of subsequent processing; Reprojection algorithm calculates the reprojection position of feature points in the world coordinate system, provides a basis for error calculation, and ensures the accuracy of measurement results; Bilinear compensation algorithm, based on the error of known points, compensates through bilinear interpolation to further improve measurement accuracy; At the same time, the following steps are implemented through the above modules: S1: Perform image acquisition, then obtain user input through the human-computer interaction module, move the motion platform to the specified position, and record the position and posture in real time; S2: The data processing module processes, corrects and compensates the acquired images; S3: Output high-precision measurement results.

2. The visual measurement system based on bilinear interpolation error compensation according to claim 1, characterized in that: The distortion correction algorithm obtains the internal and external parameters of the camera by extracting and calibrating the image feature points, and extracts and calibrates the feature points based on the image obtained by image acquisition to obtain the internal and external parameters of the camera.

3. The visual measurement system based on bilinear interpolation error compensation according to claim 1, characterized in that: The calibration coordinate system solves the matrix of the machine and the camera, places the calibration plate on the motion platform, performs translation operation, records 10-15 sets of motion coordinates RealMacPos of the machine and the corresponding calibration plate coordinates PixPos after distortion correction, and obtains the homography matrix through affine transformation.

4. The visual measurement system based on bilinear interpolation error compensation according to claim 1, characterized in that: The purpose of the reprojection algorithm is to convert the image coordinate system to the machine coordinate system by interpolating the coordinates CalMacPos and RealMacPos as reference errors, where: CalMacPos=M·PixPos.

5. The visual measurement system based on bilinear interpolation error compensation according to claim 1, characterized in that: When the bilinear interpolation compensation algorithm is used, the error values ​​in the x and y directions are used as reference values, and reverse compensation is performed to make the error tend to 0. The specific steps are as follows: Step 1: Select four nearby known measurement points and their corresponding reprojection errors v 00 ,v 01 ,v 10 ,v 11 ; Step 2: Use the bilinear interpolation formula to calculate the compensation value of the target position based on the error value of the known point. The model is as follows: Where v(x,y) represents the error value of the (x,y) position, X and Y are the lengths of the sides in the x and y directions respectively; The measurement obtains the current image position PixPos1 by image processing and calculates: MacPos=M·PixPos1 Step 3: Update the position by the compensation value to obtain the new measurement coordinates: NewMacPos / MacPos+v.

Images