A target-based machine vision displacement measurement method, system, and device

Abstract

This invention discloses a target-based machine vision displacement measurement method, system, and device, belonging to the field of machine vision displacement measurement technology. It solves the problems of current machine vision displacement measurement methods, such as the inability to automatically identify targets and the susceptibility of displacement measurement to image distortion. The key technical points are: acquiring a target image; identifying the target code based on the coded region of the target image; correcting the camera distortion and perspective imaging deformation of the target image based on the concentric multi-pattern region; calculating the centroid based on the corrected image; verifying the centroid through multiple centrally symmetric patterns to obtain the target centroid; and calculating and measuring the target displacement value based on the target centroid. Identifying the target code through the coded region and achieving data measurement and verification through the concentric multi-pattern region improves calculation accuracy.

Description

Technical Field

[0001] This invention relates to the field of machine vision displacement measurement technology, and more specifically, to a target-based machine vision displacement measurement method, system, and device. Background Technology

[0002] With the continuous development of machine vision technology and artificial intelligence algorithms, machine vision displacement measurement technology has become increasingly usable, offering advantages such as high speed and simultaneous multi-point measurement. Conventional machine vision displacement measurement involves acquiring target images, calculating the target centroid to obtain the target center coordinates, and then using these coordinates to calculate the relative positional changes between targets. However, conventional machine vision displacement measurement suffers from the following problems: it cannot automatically identify target numbers; it struggles to handle overlapping displacements of nearby targets in the image; and based on a single target pattern, centroid-based calculations are easily affected by imaging defects, leading to reduced data accuracy and data fluctuations.

[0003] Based on this, the inventors provide a target-based machine vision displacement measurement method, system, and device to solve the above problems. Summary of the Invention

[0004] The purpose of this application is to provide a target-based machine vision displacement measurement method, system, and device, solving the problems of current machine vision displacement measurement methods that cannot automatically identify targets and are easily affected by image distortion. This solution identifies target codes through coded regions and achieves image distortion repair and data verification through concentric multi-pattern regions, thereby improving calculation accuracy.

[0005] In a first aspect, this application provides a target-based machine vision displacement measurement method, comprising: a reference target placed at a fixed position and a measurement target placed at a measurement position, wherein both the reference target and the measurement target include an encoding region and a concentric multi-pattern region, the encoding region including a marker point and multiple reflective regions, and the concentric multi-pattern region including multiple concentrically arranged centrally symmetrical patterns, the reference target and the measurement target being arranged parallel to each other and placed within the visual range of a machine vision acquisition instrument; the machine vision displacement measurement method further comprises: acquiring multiple targets within the visual range when light illuminates the target. The process involves: identifying target codes based on the encoded regions of the target images; dividing multiple target images into reference target images and measurement target images based on the target codes; performing lens distortion correction on the reference target images and measurement target images; correcting perspective imaging distortion on the corrected reference target images and measurement target images based on the concentric multi-pattern regions; calculating and verifying the centroids of the corrected reference target images and measurement target images based on the concentric multi-pattern regions to obtain the centroids of the reference target and measurement target; and calculating the displacement change of the measurement target centroid relative to the reference target centroid as the measurement target displacement value.

[0006] By employing the above technical solution, target images are acquired based on machine vision, and the codes of each target are automatically identified based on these images. This enables target code recognition and group management, effectively addressing the identification and displacement measurement of targets when they are shifted in an overlapping manner. Machine vision can identify multiple concentric patterns, and deformation correction of target images and measurement value verification can be achieved through multiple concentric centrally symmetrical patterns, thereby improving the system's measurement accuracy.

[0007] In one possible implementation, identifying the target code based on the coded region of the target image includes: sequentially reading the state of each reflective region starting from the marked point, encoding the reflective region based on whether it emits light, and forming a multi-bit binary number as the target code.

[0008] In one possible implementation, the target encoding is divided into three types according to the different encoding methods: full sequence number encoding, group encoding, and check encoding; the full sequence number encoding refers to using multiple binary bits as sequence number encoding; the group encoding refers to using a portion of multiple binary bits as group number encoding and a portion as sequence number encoding; the check encoding refers to using a portion of multiple binary bits as encoding and a portion as check code.

[0009] In one possible implementation, lens distortion correction is performed on the reference target image and the measurement target image, including correcting barrel distortion and pincushion distortion in the reference target image and the measurement target image.

[0010] In one possible implementation, perspective imaging distortion correction is performed on the calibrated reference target image and the measured target image based on the concentric multi-pattern region, including: extracting multiple sets of symmetrical point coordinates from the concentric multi-pattern region of the calibrated reference target image and the measured target image; solving multiple perspective transformation matrices using the multiple sets of symmetrical point coordinates; performing outlier filtering and residual averaging on the multiple perspective transformation matrices respectively to obtain an image perspective transformation matrix applicable to both the reference target image and the measured target image; and performing distortion processing on the reference target image and the measured target image based on the image perspective transformation matrix to obtain the perspective imaging distortion corrected reference target image and the measured target image.

[0011] In one possible implementation, the centroids of the corrected reference target image and the measured target image are calculated and verified based on the concentric multi-pattern region to obtain the centroids of the reference target and the measured target. This includes: extracting multiple centrally symmetric patterns from the corrected reference target image and the measured target image respectively; calculating the coordinates of the center points of the reference target image and the measured target image respectively based on the centrally symmetric patterns; filtering outliers from the center point coordinates of the reference target image and the measured target image according to national standards; and averaging the remaining center point coordinate values ​​to obtain the centroids of the reference target and the measured target.

[0012] In a second aspect, this application provides a target-based machine vision displacement measurement system, comprising: a target and a machine vision acquisition device; the target includes a reference target placed at a fixed position and a measurement target placed at the measurement position, both the reference target and the measurement target include an encoding region and a concentric multi-pattern region, the encoding region includes a marker point and multiple reflective regions, the concentric multi-pattern region includes multiple concentrically arranged centrally symmetrical patterns, the reference target and the measurement target are arranged parallel to each other and placed within the visual range of the machine vision acquisition device; the machine vision acquisition device is used to perform a machine vision displacement measurement method as described above.

[0013] In one possible implementation, the concentric multi-pattern area includes a concentrically arranged annular pattern, a meander square pattern, and a central square pattern. The meander square pattern is arranged horizontally, and the central square pattern is arranged rotated 45 degrees. The diagonal of the central square pattern is orthogonal to the side of the meander square pattern.

[0014] In one possible implementation, the coding area includes one identification point and eight reflective areas.

[0015] In one possible implementation, the coding region is arranged around the concentric multi-pattern region.

[0016] In one possible implementation, the target is square, the coding area is placed around the target, the concentric multi-pattern area is placed inside the target, the marker point of the coding area is located at one corner of the target, and the multiple reflective areas of the coding area are grouped in pairs and placed at the four corners of the target.

[0017] In one possible implementation, both the reference target and the measurement target plane are perpendicular to the optical axis of the machine vision acquisition instrument.

[0018] In one possible implementation, the system further includes a remote information service platform for receiving the displacement value of the target calculated by the machine vision acquisition instrument.

[0019] In a third aspect, this application provides a machine vision acquisition device, comprising: an acquisition module for emitting light toward a target and acquiring multiple target images; a processor for executing a machine vision displacement measurement method as described above; a communication module for transmitting the measured target displacement values ​​to a remote information service platform; and a power supply module for supplying power to the acquisition module, the processor, and the communication module.

[0020] Compared with existing technologies, this application has the following advantages: The present invention provides a target-based machine vision displacement measurement method, system, and device. It acquires target images based on machine vision, automatically identifies the codes of each target based on the target images, and realizes target code recognition and group management, effectively handling the identification and displacement measurement of targets when adjacent targets are shifted. Machine vision can identify concentric multiple patterns, and achieves distortion correction and perspective imaging deformation correction of target images through orthogonal and symmetrical patterns. When calculating the centroid of the target, the measured values ​​can also be verified based on the concentric multiple patterns, improving the system's measurement accuracy. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0022] Figure 1 A flowchart of the target-based machine vision displacement measurement method provided by the present invention;

[0023] Figure 2 This is a schematic diagram of the perspective imaging deformation of the target image provided by the present invention;

[0024] Figure 3 This is a schematic diagram of the target-based machine vision displacement measurement system provided by the present invention.

[0025] Figure 4 A schematic diagram of the concentric multi-pattern region of the target image provided by the present invention;

[0026] Figure 5 A schematic diagram of the target image coding region provided by the present invention;

[0027] Figure 6 This is a schematic diagram of the structure of a machine vision acquisition device provided by the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the embodiments and accompanying drawings. The illustrative embodiments and descriptions of this application are only for explaining this application and are not intended to limit this application.

[0029] Current machine vision displacement measurement methods suffer from the inability to automatically identify targets, and displacement measurements are easily affected by image distortion. To address this, this application provides a target-based machine vision displacement measurement method, system, and device. It automatically obtains the target code by identifying the coded region using machine vision, and extracts multiple concentric and orthogonal patterns using machine vision to achieve centroid calculation and verification, thereby improving the accuracy of displacement measurement data calculation.

[0030] Please see Figure 1 As shown, Figure 1 This is a flowchart of a target-based machine vision displacement measurement method. A first aspect of the embodiment provides a target-based machine vision displacement measurement method. The method is based on: a reference target placed at a fixed position and a measurement target placed at the measurement position. Both the reference target and the measurement target include an encoding region and a concentric multi-pattern region. The encoding region includes a marker point and multiple reflective regions. The concentric multi-pattern region includes multiple concentrically arranged centrally symmetrical patterns. The reference target and the measurement target are arranged parallel to each other and placed within the field of view of a machine vision acquisition device. The machine vision displacement measurement method includes:

[0031] S1. When light shines on the target, acquire multiple target images within the visible range;

[0032] S2. Identify the target code based on the coding region of the target image, and divide multiple target images into reference target images and measurement target images based on the target code;

[0033] S3. Perform lens distortion correction on the reference target image and the measurement target image;

[0034] S4. Based on the concentric multi-pattern region, perform perspective imaging deformation correction on the calibrated reference target image and the measured target image;

[0035] S5. Based on the concentric multi-pattern region, calculate and verify the centroid of the corrected reference target image and the measurement target image to obtain the centroid of the reference target and the centroid of the measurement target.

[0036] S6. Calculate the displacement change of the centroid of the measuring target relative to the centroid of the reference target, and use it as the displacement value of the measuring target.

[0037] Specifically, the targets (reference target and measuring target) are hollowed out with coded areas and concentric multi-pattern areas. The hollowed-out parts of the targets are reflective films that reflect light after being illuminated by infrared LEDs. The coded areas on the targets are pre-installed with black cover films according to the coding settings. The black cover films can control whether each reflective area reflects light, so as to achieve independent coding of the targets.

[0038] Before measurement, coded targets are placed at the location where displacement needs to be measured as measurement targets. Another coded target is placed at a fixed point within the machine vision field of view as a reference target. All targets must be arranged in parallel to ensure approximate image distortion. Multiple targets are positioned within the machine vision range. During measurement, infrared LEDs illuminate the targets. The coded areas and concentric multi-pattern areas of the targets reflect light, allowing the machine vision to capture the target image. The target codes are identified, and the measurement target image and reference target image are distinguished according to the pre-recorded reference target codes and measurement target codes. Lens distortion correction and perspective distortion correction are applied to the image to obtain the corrected target image. Multiple centroid calculations are performed using the concentric multi-pattern areas, and the data from these calculations is verified to obtain the target centroid. Displacement measurement is performed based on the centroids of the measurement target and the reference target to obtain accurate measurement results.

[0039] Understandably, this method acquires target images based on machine vision, automatically identifies the codes of each target based on the images, and achieves target code recognition and group management. This effectively addresses the identification and displacement measurement of targets when they are shifted and interleaved. Machine vision can identify concentric multiple patterns, using these patterns to correct distortion and perspective imaging deformation in the target images. When calculating the target centroid, the measured values ​​are verified based on the concentric multiple patterns, improving the system's measurement accuracy.

[0040] In one possible implementation, identifying the target code based on the coded region of the target image includes: sequentially reading the state of each reflective region starting from the marked point, encoding the reflective region based on whether it emits light, and forming a multi-bit binary number as the target code.

[0041] Specifically, the state of each reflective area can be read clockwise, counterclockwise, or in a manually set specific order, starting from the marked point. The state includes whether it is luminous or non-luminous, which can be represented digitally as 0 or 1. This data is then combined to form a multi-bit binary number as the target code. For example, with 8 reflective areas, each area is categorized as reflective or non-reflective depending on whether it is obscured by a black covering film. Reflective is recorded as 1, and non-reflective as 0. During identification, starting from the marked point, the states of the reflective areas are read sequentially: code 0, code 1, code 2, code 3, code 4, code 5, code 6, and code 7, obtaining an eight-bit binary code as the target code.

[0042] In one possible implementation, the target encoding is divided into three types based on the encoding method: full sequence number encoding, group encoding, and check encoding; again, taking eight reflective areas as an example:

[0043] (1) The full sequence number encoding refers to using all 8 bits of binary as sequence number encoding, which can encode a total of 256 numbers from 0 to 255;

[0044] (2) The group coding refers to the use of part of the 8-bit binary code as group number coding and part as sequence number coding, as shown in Table 1.

[0045] Table 1 Group Coding Table

[0046] Group number digits Group number Serial number digits Serial Number Quantity 2 (encoded 0-1) 4 6 (Code 2-7) 64 3 (encoding 0-2) 8 5 (Code 3-7) 32 4 (Code 0-3) 16 4 (Code 4-7) 16 5 (Code 0-4) 32 3 (Code 5-7) 8 6 (Code 0-5) 64 2 (Code 6-7) 4

[0047] (3) The verification code refers to the part of the 8-bit binary code used as encoding (full sequence number encoding or block encoding) and the part used as a check code.

[0048] In one possible implementation, lens distortion correction is performed on the reference target image and the measurement target image, including correcting barrel distortion and pincushion distortion in the reference target image and the measurement target image.

[0049] Specifically, when setting up the targets, they are not completely perpendicular to the camera's optical axis, but rather at a certain angle. This will cause image distortion of the targets, including pincushion and barrel distortion. This can be corrected by conventional lens distortion methods (such as calibration targets). In addition, before measurement, multiple target planes should be installed in parallel, with the target planes perpendicular to the camera's optical axis, to ensure that the distortion of each target is consistent in the image and to simplify the correction process.

[0050] In one possible implementation, perspective imaging distortion correction is performed on the calibrated reference target image and the measured target image based on the concentric multi-pattern region, including: extracting multiple sets of symmetrical point coordinates from the concentric multi-pattern region of the calibrated reference target image and the measured target image; solving multiple perspective transformation matrices using the multiple sets of symmetrical point coordinates; performing outlier filtering and residual averaging on the multiple perspective transformation matrices respectively to obtain an image perspective transformation matrix applicable to both the reference target image and the measured target image; and performing distortion processing on the reference target image and the measured target image based on the image perspective transformation matrix to obtain the perspective imaging distortion corrected reference target image and the measured target image.

[0051] Please see Figure 2 As shown, Figure 2 This diagram illustrates perspective distortion in a target image. Taking a concentric multi-pattern region composed of a ring pattern, a square-shaped pattern, and a central square pattern as an example, three squares, F1, F2, and F3, are extracted from the square-shaped and central square regions. The perspective transformation matrix is ​​calculated based on the four corner points of each square. Three perspective transformation matrices are then derived for each square. Outlier values ​​are filtered out from the corresponding data points based on these three matrices, eliminating invalid data with excessive differences. The remaining valid data are averaged to obtain the final image perspective transformation matrix. The target image is then deformed based on this perspective transformation matrix to obtain the image after perspective distortion correction.

[0052] It should be noted that the above example uses a single target to illustrate the perspective imaging distortion correction process. In practical applications, there are multiple targets within the field of view. The perspective transformation matrix in all target images can be solved using the aforementioned process. After outlier filtering, the remaining valid values ​​are averaged to obtain the image perspective transformation matrix, which is used to repair the perspective imaging distortion of all target images.

[0053] In one possible implementation, the centroids of the corrected reference target image and the measured target image are calculated and verified based on the concentric multi-pattern region to obtain the centroids of the reference target and the measured target. This includes: extracting multiple centrally symmetric patterns from the corrected reference target image and the measured target image respectively; calculating the coordinates of the center points of the reference target image and the measured target image respectively based on the centrally symmetric patterns; filtering outliers from the center point coordinates of the reference target image and the measured target image according to national standards; and averaging the remaining center point coordinate values ​​to obtain the centroids of the reference target and the measured target.

[0054] Specifically, taking a concentric multi-pattern region composed of a circular pattern, a meander square pattern, and a central square pattern as an example, two circles C1 and C2 and three squares F1-F3 are extracted from the circular region, the meander square region, and the central square region. The coordinate values ​​of the center points O1 to O5 corresponding to the two circles and the three squares are calculated. Outliers of the five center point coordinate values ​​are filtered according to national standards to obtain the remaining center point data and perform average calculation to obtain the target centroid.

[0055] Finally, in step S6, the displacement change of the centroid of the measured target relative to the centroid of the reference target is calculated and used as the displacement value of the measured target.

[0056] In a second aspect, a target-based machine vision displacement measurement system is provided; please refer to [link to relevant documentation]. Figure 3 As shown, Figure 3 This is a schematic diagram of a target-based machine vision displacement measurement system. The system includes a target and a machine vision acquisition device. The target includes a reference target placed at a fixed position and a measurement target placed at the measurement position. Both the reference target and the measurement target include an encoding region and a concentric multi-pattern region. The encoding region includes a marker point and multiple reflective regions. The concentric multi-pattern region includes multiple concentrically arranged centrally symmetrical patterns. The reference target and the measurement target are arranged parallel to each other and are within the field of view of the machine vision acquisition device. The machine vision acquisition device is used to perform a machine vision displacement measurement method as described above.

[0057] Specifically, the reference target provides a reference position for displacement measurement. The measuring target moves along with the object being measured, and the machine vision acquisition instrument realizes the acquisition, correction, centroid calculation, and displacement measurement of the target image.

[0058] The measuring target is fixed at the point to be measured and moves with the point; the reference target is fixed at a fixed point within the field of view; the machine vision acquisition instrument is fixed at the reference point by a stabilizing mechanical device. The machine vision acquisition instrument acquires images of the targets, and the machine vision displacement measurement method described above is executed to calculate the centroid coordinates of the reference target and the measuring target. The difference between these coordinates is the displacement value of the measuring target, Δx and Δy.

[0059] In one possible implementation, the concentric multi-pattern area includes a concentrically arranged annular pattern, a meander square pattern, and a central square pattern. The meander square pattern is arranged horizontally, and the central square pattern is arranged rotated 45 degrees. The diagonal of the central square pattern is orthogonal to the side of the meander square pattern.

[0060] Please see Figure 4 As shown, Figure 4 This is a schematic diagram of the concentric multi-patterned regions of a target image. Specifically, the target is concentrically arranged from the outside in with: a circular pattern, a meandering square pattern, and a central square pattern. The white areas in the diagram are fitted with reflective films, which facilitates pattern extraction by machine vision.

[0061] Please see Figure 5 As shown, Figure 5 This is a schematic diagram of the target image encoding region. The encoding region includes one marker point and eight reflective areas. Starting from the marker point, the states of the eight reflective areas are read sequentially to obtain an eight-bit binary number as the target code, enabling the grouping, management, and identification of targets.

[0062] It should be noted that, Figure 5 This is merely an example that can be used for implementation; the number of reflective areas and the encoding format are not limited to this example.

[0063] In one possible implementation, the coding region is positioned around the concentric multi-pattern region. The coding region may surround the concentric multi-pattern region from the outside or from the inside. The aim is to keep the coding region compact with the concentric multi-pattern region, minimizing the target area while maintaining functionality.

[0064] In one possible implementation, to further reduce the target area, a compact layout is adopted. The target is square, the coded area is placed on the periphery of the target, the concentric multi-pattern area is placed inside the target, the marker point of the coded area is located at one corner of the target, and the multiple reflective areas of the coded area are grouped in pairs and placed at the four corners of the target.

[0065] In one possible implementation, both the reference target and the measurement target plane are perpendicular to the optical axis of the machine vision acquisition instrument.

[0066] Specifically, in practical applications, if the target rotation angle is too large, the machine vision acquisition instrument may not be able to acquire a complete target image. Therefore, it is advisable to select that both the reference target and the measurement target are parallel to the acquisition surface of the machine vision acquisition instrument to ensure the integrity of the pattern.

[0067] In one possible implementation, the system further includes a remote information service platform for receiving the displacement value of the target calculated by the machine vision acquisition instrument.

[0068] Specifically, the target marks the reference points and measurement points, and the machine vision acquisition instrument completes the functions of on-site image acquisition, image analysis, displacement value calculation, and data and image storage. Finally, the displacement value is uploaded to the remote information service platform via the network. The remote information service platform realizes the remote configuration of the machine vision acquisition instrument's parameters, target encoding settings, data reception and display, and remote image retrieval, enabling remote measurement and management.

[0069] In a third aspect of the embodiments, a machine vision acquisition device is provided; please refer to [link to relevant documentation]. Figure 6 As shown, Figure 6 This is a schematic diagram of the structure of a machine vision acquisition device. It includes: an acquisition module for emitting light towards a target and acquiring multiple target images; a processor for executing one of the machine vision displacement measurement methods described above; a communication module for transmitting the measured target displacement values ​​to a remote information service platform; and a power supply module for supplying power to the acquisition module, processor, and communication module.

[0070] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A target-based machine vision displacement measurement method, characterized in that, include: A reference target is placed in a fixed position and a measuring target is placed in the position to be measured. Both the reference target and the measuring target include an encoding area and a concentric multi-pattern area. The encoding area includes a marker point and multiple reflective areas. The concentric multi-pattern area includes multiple concentrically set centrally symmetrical patterns. The reference target and the measuring target are arranged in parallel and placed within the field of view of the machine vision acquisition instrument. The machine vision displacement measurement method includes: Multiple target images are captured within the visible range when light illuminates the target; Target codes are identified based on the coding regions of target images, and multiple target images are divided into reference target images and measurement target images based on the target codes. Lens distortion correction is performed on the reference target image and the measured target image; Based on the concentric multi-pattern region, perspective imaging distortion correction is performed on the calibrated reference target image and the measured target image. This correction includes: extracting multiple sets of symmetrical point coordinates from the concentric multi-pattern region of the calibrated reference target image and the measured target image; solving multiple perspective transformation matrices using these coordinates; performing outlier filtering and residual averaging on the multiple perspective transformation matrices to obtain image perspective transformation matrices applicable to both the reference target image and the measured target image; and performing distortion processing on the reference target image and the measured target image based on the image perspective transformation matrices to obtain the perspective-corrected reference target image and the measured target image. Based on the concentric multi-pattern region, the centroids of the corrected reference target image and the measured target image are calculated and verified to obtain the centroids of the reference target and the measured target. The displacement change of the centroid of the measured target relative to the centroid of the reference target is calculated and used as the displacement value of the measured target.

2. The method according to claim 1, characterized in that, Identifying the target code based on the coding region of the target image includes: sequentially reading the state of each reflective area starting from the marked point, encoding the reflective area based on whether it emits light, and forming a multi-bit binary number as the target code.

3. The method according to claim 2, characterized in that, Based on different encoding methods, the target encoding is divided into: full sequence number encoding, block encoding, and check encoding; The full sequence number encoding refers to using multiple binary bits as sequence number encoding; The block coding refers to using a portion of a multi-bit binary number as a group number encoding and a portion as a sequence number encoding. The check code refers to a combination of multiple binary bits, where one part is used as the encoding and the other part as the check code.

4. The method according to claim 1, characterized in that, Lens distortion correction is performed on the reference target image and the measured target image, including: correcting barrel distortion and pincushion distortion in the reference target image and the measured target image.

5. The method according to claim 1, characterized in that, Based on the concentric multi-pattern region, the centroids of the corrected reference target image and the measured target image are calculated and verified to obtain the centroids of the reference target and the measured target, including: Multiple centrally symmetric patterns were extracted from the corrected reference target image and the measured target image, respectively; Calculate the center point coordinates of the reference target image and the measured target image based on the centrally symmetric pattern; Outlier values ​​were filtered out from the center point coordinates of the reference target image and the measured target image according to the image standard. The remaining center point coordinate values ​​were averaged to obtain the centroid of the reference target and the centroid of the measured target.

6. A target-based machine vision displacement measurement system, characterized in that, include: Targets and machine vision acquisition devices; The target includes a reference target placed at a fixed position and a measurement target placed at the position to be measured. Both the reference target and the measurement target include an encoding area and a concentric multi-pattern area. The encoding area includes a marker point and multiple reflective areas. The concentric multi-pattern area includes multiple concentrically set centrally symmetrical patterns. The reference target and the measurement target are arranged in parallel and placed within the field of view of the machine vision acquisition instrument. The machine vision acquisition device is used to perform a machine vision displacement measurement method as described in any one of claims 1-5.

7. The system according to claim 6, characterized in that, The concentric multi-pattern area includes a concentrically arranged circular pattern, a meandering square pattern, and a central square pattern. The meandering square pattern is arranged horizontally, and the central square pattern is arranged rotated 45 degrees. The diagonal of the central square pattern is orthogonal to the side of the meandering square pattern.

8. The system according to claim 6, characterized in that, The coding area includes one identification point and eight reflective areas.

9. The system according to claim 6, characterized in that, The encoding region is arranged around the concentric multi-pattern region.

10. The system according to claim 6, characterized in that, The target is square, the coding area is placed on the periphery of the target, the concentric multi-pattern area is placed inside the target, the marker point of the coding area is located at one corner of the target, and the multiple reflective areas of the coding area are grouped in pairs and placed at the four corners of the target.

11. The system according to claim 6, characterized in that, The planes of both the reference target and the measurement target are perpendicular to the optical axis of the machine vision acquisition instrument.

12. The system according to claim 6, characterized in that, The system also includes a remote information service platform, which receives the target displacement values ​​calculated by the machine vision acquisition instrument.

13. A machine vision acquisition device, comprising: The acquisition module is used to emit light towards the target and acquire multiple target images; A processor for executing a machine vision displacement measurement method as described in any one of claims 1-5; A communication module is used to transmit the displacement values ​​of the measured target to a remote information service platform; and, A power supply module is used to supply power to the acquisition module, processor, and communication module.

Images