A method and system for intelligent adjustment of markers and image acquisition control

Abstract

The present invention relates to the technical field of image analysis and acquisition, and specifically to a method and system for intelligently adjusting a marker and controlling image acquisition. The control method places a bottle of wine to be inspected on a liftable storage platform and then moves the bottle to a preset height, and based on the position matching of an acquisition window, a glass plate, a plane mirror and a camera and the construction of a soft light environment, makes the marker image acquired by the camera clearer. At the same time, when there is no complete bottleneck marker in the initial acquisition image, it is necessary to determine the position of the marker on the bottleneck, and then rotate the bottle of wine to be inspected so that the bottleneck marker faces the acquisition window to ensure that the acquired image contains a complete bottleneck marker image, thereby improving the acquisition quality of the bottleneck marker image. The method is applied in the field of wine bottle marker picture acquisition, and can not only realize automatic image acquisition control, but also acquire complete and high-quality marker pictures, which is conducive to enhancing the accuracy and reliability of product recognition results.

Description

Technical Field

[0001] The present invention relates to the general field of control technology, and in particular to a method and system for intelligently adjusting a marker and controlling image acquisition. Background Art

[0002] The identification on the wine bottle, such as the QR code on the neck of the bottle, records the production information of the wine corresponding to the bottle. By identifying the identification code, the entire production and transportation process of the wine can be traced to ensure product quality and authenticity.

[0003] Currently, the quality of the pictures of wine bottle markers is often controlled by manually controlling the positions of the wine bottles and cameras. However, manual photography requires operators to take pictures of the wine bottles one by one, which is relatively slow. Especially in large-scale production or testing environments, the problem of low efficiency is more prominent, which will increase the time cost of production and testing and reduce overall work efficiency. In addition, different operators may have differences when taking pictures, resulting in inconsistent image quality and angles, which will bring difficulties to subsequent marker recognition and wine information analysis, affecting the accuracy and reliability of product recognition results. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for intelligently adjusting a marker and controlling image acquisition.

[0005] The technical solution of the present invention is as follows:

[0006] A method for intelligently adjusting a marker and controlling image acquisition includes the following operations:

[0007] S1. Put the bottled wine to be inspected on a liftable storage platform, which drives the bottled wine to be inspected to move upward to a preset bottle height. A soft light cover is provided at the preset bottle height, with the soft light cover opening downward. A collection window is provided on the soft light cover, a plane mirror is obliquely provided above one side of the collection window, a glass plate is vertically provided between the collection window and the plane mirror, and a first camera is provided below the plane mirror. The positional relationship between the collection window, the plane mirror, the glass plate, and the first camera can satisfy the requirement that the first camera can capture the image of the bottled wine to be inspected on the plane mirror after passing through the collection window and the glass plate after fill light, so as to obtain an initial collection image.

[0008] S2, determine whether there is a complete bottleneck marker in the initial collection map; if so, use the initial collection map as the bottleneck marker collection map; if not, execute S3;

[0009] S3. Rotate the bottled wine to be inspected on the liftable storage table so that the second camera can obtain an image of the bottleneck of the bottled wine to be inspected; after splicing the images of the bottleneck, a bottleneck splicing image is obtained; the installation height of the second camera is a preset wine bottle height; based on the position of the bottleneck marker in the bottleneck splicing image, as well as the positions of the second camera and the acquisition window, the adjustment direction and adjustment angle of the bottled wine to be inspected are obtained; according to the adjustment direction and adjustment angle of the bottled wine to be inspected, the bottled wine to be inspected is rotated until the bottleneck marker is facing the acquisition window, and the first camera captures the image on the plane mirror to obtain a bottleneck marker acquisition map.

[0010] The operation of obtaining the bottleneck stitching image in S3 is specifically as follows: based on the viewing angle of the second camera, determine the single rotation angle of the bottled wine to be inspected and the total number of rotations; rotate the bottled wine to be inspected several times according to the single rotation angle and the total number of rotations, and after each rotation stops, obtain the bottleneck image of the bottled wine to be inspected by the second camera until the bottled wine to be inspected rotates one circle, and obtain several bottleneck images; among the several bottleneck images, the bottleneck images with the first and second image acquisition times are subjected to image fusion processing based on matching points at the stitching area to obtain a first stitching image; repeat the image fusion processing based on matching points at the stitching area, and stitch the already stitched image with the next bottleneck image until all the bottleneck images are stitched together to obtain the bottleneck stitching image.

[0011] The operations of performing matching point-based image fusion processing on the bottleneck images whose image acquisition time is the first and the second are as follows: taking a position point whose contrast is greater than a contrast threshold in a neighborhood range at a stitching area in the bottleneck image whose image acquisition time is the first as a first matching point; taking a position point corresponding to the first matching point in the stitching area of ​​the bottleneck image whose image acquisition time is the second as a second matching point; calculating a transformation matrix based on all the first matching points and the second matching points; and performing pixel-weighted fusion processing on the stitching area in the bottleneck images whose image acquisition time is the first and the second based on the transformation matrix to obtain the first stitching image.

[0012] The method for identifying bottleneck markers in S2 is specifically as follows: the bottleneck stitching image is covered with detection frames, and the detection frame corresponding to the minimum distance between the center point of the detection frame and the center point of the real frame in the same area is used as the frame to be classified in the corresponding area; all the frames to be classified are classified to obtain the bottleneck marker recognition result.

[0013] In S2, before identifying the bottleneck marker, it also includes performing image enhancement processing on the bottleneck stitching image to obtain a bottleneck stitching enhanced image for performing the identification operation; the image enhancement processing operation is specifically as follows: if the information entropy of the bottleneck stitching image is greater than the information entropy threshold, the bottleneck stitching image is divided into a number of sub-regions, and each sub-region is subjected to histogram equalization processing according to the grayscale distribution information of each sub-region until the contrast of each sub-region reaches the corresponding contrast threshold, and then stitching is performed to obtain a bottleneck stitching shallow enhanced image; the pixel value of each position point at the stitching boundary in the bottleneck stitching shallow enhanced image is changed to the average pixel value of the position points within the corresponding four neighborhoods to obtain a bottleneck stitching enhanced image; if the information entropy of the bottleneck stitching image is not greater than the information entropy threshold, the bottleneck stitching image is sequentially subjected to closing operation and opening operation processing to obtain a bottleneck stitching shallow enhanced image; after obtaining the difference map between the bottleneck stitching shallow enhanced image and the bottleneck stitching image, it is superimposed with the bottleneck stitching image to obtain a bottleneck stitching enhanced image.

[0014] A marker intelligent adjustment and image acquisition control system, used to implement the above-mentioned marker intelligent adjustment and image acquisition control method, comprising:

[0015] A liftable storage table is used to place the bottled wine to be inspected and drive the bottled wine to be inspected to move;

[0016] The collection module includes a soft light cover, a collection window, a glass plate, a plane mirror, and a first camera; the height of the soft light cover is a preset wine bottle height, the soft light cover opens downward, the collection window is located on the soft light cover, a plane mirror is obliquely arranged above one side of the collection window, a glass plate is vertically arranged between the collection window and the plane mirror, and a first camera is arranged below the plane mirror; the positional relationship of the collection window, the plane mirror, the glass plate, and the first camera can satisfy the requirement that the first camera can capture the image of the bottled wine to be inspected on the plane mirror passing through the collection window and the glass plate after supplementary light, obtain an initial collection image or a bottleneck marker collection image, and transmit the initial collection image to the processing module;

[0017] The processing module is used to determine whether there is a complete bottleneck marker in the initial acquisition image after receiving the initial acquisition image from the acquisition module; if so, use the initial acquisition image as the bottleneck marker acquisition image; if not, transmit a signal indicating that there is no complete bottleneck marker in the initial acquisition image to the control module; and after receiving the bottleneck circle image from the second camera, splice the bottleneck circle image to obtain a bottleneck splicing image; based on the position of the bottleneck marker in the bottleneck splicing image, as well as the position of the second camera and the acquisition window, obtain the adjustment direction and adjustment angle of the bottled wine to be inspected, and transmit the signal to the control module;

[0018] The control module is used to control the up and down movement of the liftable storage platform, and when receiving a signal from the processing module that there is no complete bottleneck marker in the initial acquisition image, control the rotation of the bottled wine to be inspected, so that the second camera obtains a circle image of the bottleneck of the bottled wine to be inspected, and the installation height of the second camera is a preset wine bottle height; and according to the direction and angle of the bottled wine to be inspected transmitted by the processing module, the bottled wine to be inspected is rotated until the bottleneck marker is facing the acquisition window, and then the first camera is controlled to take a picture.

[0019] A first supplementary light source is obliquely arranged on both sides above the soft light cover, and the illumination direction of the first supplementary light source is directly facing the side wall of the soft light cover; a vertical second supplementary light source is arranged between the collection window and the glass plate.

[0020] When the bottled wine to be inspected rises to the preset bottle height, the top height of the bottleneck of the bottled wine to be inspected does not exceed the top height of the collection window, and the bottom height of the bottleneck of the bottled wine to be inspected does not fall below the bottom height of the collection window.

[0021] The beneficial effects of the present invention are:

[0022] The present invention provides a method for intelligently adjusting and controlling an image acquisition of a marker. The method places a bottle of wine to be inspected on a liftable storage platform and moves the bottle to a preset height. The method also matches the positions of an acquisition window, a glass plate and a plane mirror and constructs a soft light field environment to reduce the influence of uneven light on a camera, thereby making the marker image acquired by the camera clearer and more realistic. At the same time, in this process, the method also determines whether to perform an operation of adjusting the bottle of wine to be inspected based on the integrity of the bottleneck marker in the initial acquisition image. When a complete bottleneck marker does not exist in the initial acquisition image, the position of the marker on the bottleneck needs to be determined. Then, by rotating the bottle of wine to be inspected, the bottleneck marker is aligned with the acquisition window to ensure that the acquired image contains a complete bottleneck marker image, thereby improving the acquisition quality of the bottleneck marker image. The method is applied in the field of wine bottle marker picture acquisition, and can not only realize automated image acquisition control, but also acquire complete and high-quality marker pictures, which is beneficial to enhancing the accuracy and reliability of product recognition results. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] By reading the detailed description of the preferred embodiment below, the scheme and advantages of the present application will become clear to those skilled in the art. The accompanying drawings are only for the purpose of illustrating the preferred embodiment and are not to be considered as limiting the present invention.

[0024] In the attached picture:

[0025] Figure 1 It is a structural schematic diagram of the acquisition control system in the embodiment;

[0026] The components represented by the reference numerals in the figure are:

[0027] A liftable storage table; 2. a rotating tray; 3. a soft light cover; 4. a glass plate; 5. a plane mirror; 6. a first camera; 7. a first supplementary light source; 8. a second supplementary light source; 9. a first photoelectric sensor; 10. a second photoelectric sensor. DETAILED DESCRIPTION

[0028] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings.

[0029] This embodiment provides a marker intelligent adjustment and image acquisition control system, see Figure 1 Some components in are used to realize the intelligent adjustment of markers and the image acquisition control method, including:

[0030] The liftable storage table 1 is used to place the bottled wine to be inspected and drive the bottled wine to be inspected to move; preferably, a rotating tray 2 is provided on the liftable storage table, and when receiving the rotation signal of the control module, the rotating tray 2 will drive the bottled wine to be inspected to rotate.

[0031] The acquisition module includes a soft light cover 3, a collection window, a glass plate 4, a plane mirror 5, and a first camera 6; the height of the soft light cover 3 is a preset wine bottle height, the soft light cover 3 opens downward, the acquisition window is located on the soft light cover 3, a plane mirror 5 is obliquely arranged above one side of the collection window, a glass plate 4 is vertically arranged between the collection window and the plane mirror 5, and a first camera 6 is arranged below the plane mirror 5; the positional relationship of the collection window, the plane mirror 5, the glass plate 4, and the first camera 6 can satisfy that the first camera 6 can capture the image of the bottled wine to be inspected on the plane mirror 5 after passing through the collection window and the glass plate 4 after fill light, to obtain an initial acquisition image or a bottleneck marker acquisition image, and transmit the initial acquisition image to the processing module.

[0032] When the bottled wine to be inspected rises to a preset bottle height, it enters the soft light cover 3, and the bottleneck marker of the bottled wine to be inspected is mapped onto the glass plate 4 through the acquisition window. The first camera 6 uses the shooting plane mirror 5 to image the bottleneck marker on the glass plate 4 to obtain an initial acquisition image or a bottleneck marker acquisition image.

[0033] Among them, the size of the collection window is larger than the size of the marker, and when the bottle of wine to be inspected rises to the preset bottle height, the top height of the bottleneck of the bottle of wine to be inspected does not exceed the top height of the collection window, and the bottom height of the bottleneck of the bottle of wine to be inspected is not lower than the bottom height of the collection window.

[0034] To ensure that the bottleneck marker of the bottled wine to be inspected can be obtained, the present embodiment arranges that when the bottled wine to be inspected rises to a preset bottle height, the top height of the bottleneck of the bottled wine to be inspected does not exceed the top height of the collection window on one side of the bottled wine to be inspected, and the bottom height of the bottleneck of the bottled wine to be inspected is not lower than the bottom height of the collection window.

[0035] At the same time, a first photoelectric sensor 9 is provided on one side of the outside of the soft light cover 3. The installation height of the first photoelectric sensor 9 is the preset wine bottle height, and is used to detect whether the bottled wine to be inspected moves to the preset wine bottle height.

[0036] Furthermore, a second photoelectric sensor 10 is provided on the upper part of the first photoelectric sensor 9. In the process of the liftable storage platform driving the bottled wine to be inspected to move upward, if the liftable storage platform does not stop moving after the bottled wine to be inspected rises to the preset bottle height, that is, the second photoelectric sensor 10 above the first photoelectric sensor 9 detects that the top of the bottled wine to be inspected exceeds the safe height, the second photoelectric sensor 10 sends an alarm signal to the control module. After receiving the alarm signal, the control module controls the liftable storage platform 1 to move downward until the bottled wine to be inspected drops to the preset bottle height.

[0037] In addition, a first supplementary light source 7 is obliquely provided on both sides above the soft light cover 3, and the lighting direction of the first supplementary light source 7 is directly facing the side wall of the soft light cover 3, so as to avoid the first supplementary light source 7 directly facing the bottled wine to be inspected and causing overexposure of the collected image, and provide a soft light environment for image collection, which is beneficial to improving the image collection quality; and a vertical second supplementary light source 8 is provided between the collection window and the glass plate 4, or the glass plate 4 is embedded in the light source plate containing the second supplementary light source 8, so as to prevent the problem of low brightness and blur in the middle of the marker image caused by insufficient lighting of the middle part of the bottleneck by the first supplementary light source 7, and also enhance the brightness near the glass plate 4, so that the image received by the plane mirror 5 is clearer and the image quality collected by the first camera 6 is better.

[0038] The processing module is used to determine whether there is a complete bottleneck marker in the initial acquisition image after receiving the initial acquisition image from the acquisition module; if so, use the initial acquisition image as the bottleneck marker acquisition image to obtain the information of the bottled wine to be inspected; if not, transmit a signal indicating that there is no complete bottleneck marker in the initial acquisition image to the control module; and after receiving the bottleneck circle image from the second camera, splice the bottleneck circle image to obtain a bottleneck splicing image; based on the position of the bottleneck marker in the bottleneck splicing image, as well as the position of the second camera and the acquisition window, obtain the adjustment direction and adjustment angle of the bottled wine to be inspected, and transmit the signal to the control module;

[0039] The control module is used to control the liftable storage platform 1 to move up and down, and when receiving a signal from the processing module that there is no complete bottleneck marker in the initial acquisition image, control the rotation of the bottled wine to be inspected, so that the second camera obtains a circle image of the bottleneck of the bottled wine to be inspected, and the installation height of the second camera is a preset wine bottle height; and according to the direction and angle of the bottled wine to be inspected transmitted by the processing module, the bottled wine to be inspected is rotated until the bottleneck marker is facing the acquisition window, and then the first camera is controlled to take a picture.

[0040] This embodiment also provides a method for intelligent adjustment of a marker and image acquisition control, and the specific steps are as follows.

[0041] S1. Put the bottled wine to be inspected on a liftable storage platform. The liftable storage platform drives the bottled wine to be inspected to move upward to a preset bottle height. A soft light cover is provided at the preset bottle height, and the soft light cover opens downward. A collection window is provided on the soft light cover. A plane mirror is obliquely provided above one side of the collection window. A glass plate is vertically provided between the collection window and the plane mirror. A first camera is provided below the plane mirror. The positional relationship among the collection window, the plane mirror, the glass plate and the first camera can satisfy the requirement that the first camera can capture the image of the bottled wine to be inspected on the plane mirror after passing through the collection window and the glass plate after fill light, so as to obtain an initial collection image.

[0042] After placing the bottled wine to be inspected on a liftable storage table, the bottled wine to be inspected is moved to a preset bottle height and enters a soft light cover with a soft light environment to prevent the surface of the bottled wine to be inspected from reflecting too brightly. Based on the position matching of the collection window, glass plate and plane mirror on the soft light cover, the influence of uneven light on the camera can be reduced, so that the image obtained by the camera is clearer and more realistic, and the image collection quality of the bottleneck marker is improved.

[0043] First, place the bottled wine to be inspected on a liftable storage table, specifically on a rotating tray on the upper part of the liftable storage table. The telescopic rod on one side of the liftable storage table drives the liftable storage table to move upward. At this time, the liftable storage table drives the bottled wine to be inspected to move upward. When the first photoelectric sensor on the outer side of the soft light cover recognizes that the bottled wine to be inspected has risen to a preset bottle height, the first photoelectric sensor sends a signal to the control module. After receiving the signal, the control module controls the liftable storage table to stop moving.

[0044] Then, a collection window is opened on the inner wall of the soft light cover at a preset height of the wine bottle. The collection window of fixed size can physically block the contents of the bottle except the bottleneck, ensuring that the marker on the bottleneck can be identified as soon as possible in the acquired image; at the same time, a plane mirror is obliquely arranged above the other side of the collection window (45° tilted downward), and a glass plate is vertically arranged between the collection window and the plane mirror. The glass plate not only facilitates the plane mirror to obtain a complete image of the bottleneck marker, but also can filter out a part of the stray reflected light, making the image received by the plane mirror clearer.

[0045] Finally, a first camera is provided below the plane mirror, and the positional relationship among the acquisition window, plane mirror, glass plate and the first camera is such that the first camera can capture the image of the bottled wine to be inspected on the plane mirror passing through the acquisition window and glass plate to obtain an initial acquisition image.

[0046] S2. Determine whether there is a complete bottleneck marker in the initial collection map; if so, use the initial collection map as the bottleneck marker collection map to obtain the bottled wine information to be inspected; if not, execute S3.

[0047] According to the integrity of the bottleneck marker in the initial acquisition image, choose whether to perform the operation of adjusting the bottled wine to be inspected. On the basis of ensuring the image quality of the bottleneck marker, it can reduce duplication of work and improve acquisition efficiency.

[0048] To ensure that the collected image contains complete bottleneck markers, after the initial collection image is obtained for the first time, it is necessary to determine whether there is a complete bottleneck marker in the initial collection image, that is, to determine whether the structural similarity between the bottleneck marker and the standard bottleneck marker in the initial collection image is greater than the similarity threshold.

[0049] If the structural similarity is greater than the similarity threshold, whether there is a complete bottleneck marker in the initial collection image, the initial collection image can be directly used as the bottleneck marker collection image to obtain the bottled wine information to be inspected.

[0050] If the structural similarity is not greater than the similarity threshold, then there is no complete bottleneck marker in the initial acquisition image, and it is necessary to perform the operations of adjusting the bottled wine to be inspected and re-acquiring the bottleneck marker image in S3.

[0051] The above-mentioned method for identifying bottleneck markers is specifically as follows: the initial acquisition image is covered with detection frames, and the detection frame corresponding to the minimum distance between the center point of the detection frame and the center point of the real frame in the same area is used as the frame to be classified in the corresponding area; all the frames to be classified are classified to obtain the bottleneck marker recognition result.

[0052] In order to improve the accuracy of recognition, before identifying the bottleneck marker, the initial acquisition image is also subjected to image enhancement processing to obtain an initial acquisition enhanced image for performing the recognition operation. The image enhancement processing operation is the same as that in S3. To save space, the specific process of the image enhancement processing is placed in S3 for discussion.

[0053] S3. Rotate the bottled wine to be inspected on the liftable storage table so that the second camera can obtain an image of the bottleneck of the bottled wine to be inspected; after splicing the images of the bottleneck, a bottleneck splicing image is obtained; the installation height of the second camera is a preset wine bottle height; based on the position of the bottleneck marker in the bottleneck splicing image, as well as the positions of the second camera and the acquisition window, the adjustment direction and adjustment angle of the bottled wine to be inspected are obtained; according to the adjustment direction and adjustment angle of the bottled wine to be inspected, the bottled wine to be inspected is rotated until the bottleneck marker is facing the acquisition window, and the first camera captures the image on the plane mirror to obtain a bottleneck marker acquisition map.

[0054] When there is no complete bottleneck marker in the initial acquisition image, it is necessary to determine the position of the marker on the bottleneck, and then rotate the liftable storage table or rotating tray to align the bottleneck marker on the bottled wine to be inspected with the acquisition window to ensure that the acquired image contains a complete bottleneck marker image, thereby improving the image acquisition quality.

[0055] First, the bottled wine to be inspected is rotated so that the second camera can obtain a circle image of the neck of the bottled wine to be inspected, and the circle image of the neck is spliced ​​to obtain a neck spliced ​​image. The second camera can be set on both sides of the acquisition window, or on the opposite side or on the same side, and the installation height of the second camera is the preset bottle height.

[0056] The specific operation of obtaining the bottleneck stitching image is as follows: based on the viewing angle of the second camera (obtained based on the focal length of the second camera), determine the single rotation angle of the bottled wine to be inspected and the total number of rotations. According to the single rotation angle and the total number of rotations, rotate the bottled wine to be inspected several times. After each rotation stops, the second camera obtains the bottleneck image of the bottled wine to be inspected until the bottled wine to be inspected rotates one circle, and obtains several bottleneck images; among the several bottleneck images, the bottleneck images with the first and second image acquisition times are subjected to image fusion processing based on matching points at the stitching area to obtain the first stitching image; repeat the image fusion processing based on matching points at the stitching area, and stitch the already stitched image with the next bottleneck image until all the bottleneck images are stitched together to obtain the bottleneck stitching image. The stitching area is the area with the same elements in the two images, which can be set according to actual needs.

[0057] The above-mentioned operation of performing image fusion processing based on matching points on the bottleneck images with the first and second image acquisition times is specifically as follows: taking the position points (position points with the same elements) with a contrast greater than a contrast threshold in the neighborhood range of the stitching area in the bottleneck image with the first image acquisition time as the first matching point; taking the corresponding position points of the first matching point in the stitching area of ​​the bottleneck image with the second image acquisition time as the second matching point; calculating the transformation matrix based on all the first matching points and the second matching points; based on the transformation matrix, fusing the stitching area in the bottleneck images with the first and second image acquisition times to obtain an initial stitching image; performing pixel weighted fusion processing on the pixel points at the stitching area in the initial stitching image to make the transition at the stitching area natural, and obtaining the first stitching image.

[0058] The first stitched image and the bottleneck image acquired at the third time are subjected to image fusion processing based on matching points to obtain the first stitched image; and similarly, the stitched image is stitched with the bottleneck image acquired at the last time to obtain the bottleneck stitched image.

[0059] Then, based on the position of the bottleneck marker in the bottleneck stitching image, as well as the positions of the second camera and the acquisition window, the adjustment direction and adjustment angle of the bottled wine to be inspected are obtained.

[0060] The specific method for identifying bottleneck markers is as follows: the bottleneck stitching image is covered with detection frames, and the detection frame corresponding to the minimum distance between the center point of the detection frame and the center point of the real frame in the same area is used as the frame to be classified in the corresponding area; all the frames to be classified are classified to obtain the bottleneck marker recognition result.

[0061] In order to improve the accuracy of recognition, before recognizing the bottleneck marker, the bottleneck splicing image is also enhanced to obtain a bottleneck splicing enhanced image for performing the recognition operation.

[0062] The operations of image enhancement processing are as follows.

[0063] If the information entropy of the bottleneck stitching image is greater than the information entropy threshold, it means that the bottleneck circle image is rich in content and complex in texture. Therefore, the bottleneck stitching image needs to be divided into several sub-regions. According to the grayscale distribution information of each sub-region, each sub-region is subjected to histogram equalization processing to enhance the local contrast and image details. After the contrast of each sub-region reaches the corresponding contrast threshold, the sub-regions are stitched together to obtain a bottleneck stitching shallowly enhanced image. The pixel value of each position point at the stitching boundary in the bottleneck stitching shallowly enhanced image is changed to the average pixel value of the position point within the corresponding four neighborhoods to make the transition between sub-regions smoother and avoid obvious boundary effects, so as to obtain a bottleneck stitching enhanced image.

[0064] If the information entropy of the bottleneck stitching image is not greater than the information entropy threshold, it means that the content of the bottleneck circle image is simple, and there are only markers and background. Therefore, in order to improve efficiency, the bottleneck stitching image is processed by closing operation and opening operation in sequence to obtain a bottleneck stitching shallow enhanced image; after obtaining the difference map between the bottleneck stitching shallow enhanced image and the bottleneck stitching image, it is superimposed with the bottleneck stitching image to enhance the morphological characteristics of the marker and obtain the bottleneck stitching enhanced image.

[0065] Finally, according to the adjustment direction and angle of the bottled wine to be inspected, the bottled wine to be inspected is rotated until the bottleneck marker is facing the collection window. The first camera captures the image on the plane mirror to obtain the bottleneck marker collection map for obtaining the information of the bottled wine to be inspected.

[0066] The present embodiment provides a method for intelligently adjusting a marker and controlling image acquisition. The method places a bottle of wine to be inspected on a liftable storage platform and moves it to a preset bottle height. The method also matches the positions of the acquisition window, glass plate and plane mirror and constructs a soft light field environment to reduce the impact of uneven light on the camera, thereby making the marker image acquired by the camera clearer and more realistic. At the same time, in this process, the method will also determine whether to perform the operation of adjusting the bottle of wine to be inspected based on the integrity of the bottleneck marker in the initial acquisition image. When there is no complete bottleneck marker in the initial acquisition image, it is necessary to determine the position of the marker on the bottleneck, and then rotate the bottle of wine to be inspected so that the bottleneck marker faces the acquisition window to ensure that the acquired image contains a complete bottleneck marker image, thereby improving the acquisition quality of the bottleneck marker image. The method is applied in the field of wine bottle marker image acquisition, which can not only realize automatic image acquisition, but also obtain complete and high-quality marker images, which is conducive to enhancing the accuracy and reliability of product recognition results.

Claims

1. A method for intelligent adjustment of markers and image acquisition control, characterized in that: The following operations are included: S1. Place the bottled wine to be inspected on a liftable storage platform. The liftable storage platform drives the bottled wine to be inspected to move upward to a preset bottle height. A soft light cover is provided at the preset bottle height, and the soft light cover opens downward. A collection window is provided on the soft light cover, a plane mirror is obliquely provided on one side of the collection window, a glass plate is vertically provided between the collection window and the plane mirror, and a first camera is provided below the plane mirror; the positional relationship of the collection window, the plane mirror, the glass plate, and the first camera can satisfy the requirement that the first camera can capture the image of the bottled wine to be inspected on the plane mirror passing through the collection window and the glass plate after supplementary light, so as to obtain an initial collection image; S2, determining whether there is a complete bottleneck marker in the initial collection map; If it exists, the initial collection map is used as the bottleneck marker collection map; If it does not exist, execute S3; S3, rotating the bottled wine to be inspected on the liftable storage table, so that the second camera obtains a circle image of the bottleneck of the bottled wine to be inspected; after splicing the circle image of the bottleneck, a bottleneck splicing image is obtained; the second camera is installed at a preset height of the wine bottle; Based on the position of the bottleneck marker in the bottleneck stitching image, as well as the positions of the second camera and the acquisition window, the adjustment direction and adjustment angle of the bottled wine to be inspected are obtained; According to the adjustment direction and angle of the bottled wine to be inspected, the bottled wine to be inspected is rotated until the bottleneck marker is directly opposite the collection window, and the first camera captures the image on the plane mirror to obtain a bottleneck marker collection map.

2. The method for intelligent adjustment and image acquisition control of a marker according to claim 1, characterized in that: In S3, the operation of obtaining the bottleneck stitching image is specifically as follows: Based on the viewing angle of the second camera, the single rotation angle and the total number of rotations of the bottled wine to be inspected are determined; based on the single rotation angle and the total number of rotations, the bottled wine to be inspected is rotated several times, and after each rotation stops, the second camera obtains the bottleneck image of the bottled wine to be inspected, until the bottled wine to be inspected rotates one circle, and several bottleneck images are obtained; For the bottleneck images whose image acquisition time is the first and the second among several bottleneck images, image fusion processing based on matching points is performed at the stitching area to obtain a first stitching image; the image fusion processing based on matching points at the stitching area is repeated to stitch the already stitched image with the next bottleneck image until all the bottleneck images are stitched together to obtain the bottleneck stitching image.

3. The method for intelligent adjustment and image acquisition control of a marker according to claim 2, characterized in that: The specific operation of performing image fusion processing based on matching points on the bottleneck images whose image acquisition time is the first and second is as follows: In the bottleneck image acquired at the first image acquisition time, a position point in the neighborhood range of the stitching area where the contrast is greater than the contrast threshold is used as a first matching point; a position point corresponding to the first matching point in the stitching area of ​​the bottleneck image acquired at the second image acquisition time is used as a second matching point; and a transformation matrix is ​​calculated based on all the first matching points and the second matching points; Based on the transformation matrix, pixel weighted fusion processing is performed on the stitching area in the bottleneck images with the first and second image acquisition times to obtain the first stitching image.

4. The method for intelligent adjustment of markers and image acquisition control according to claim 1, characterized in that: In S2, the method for identifying the bottleneck marker is specifically: covering the bottleneck stitching image with detection frames, and taking the detection frame corresponding to the minimum distance between the center point of the detection frame and the center point of the real frame in the same area as the to-be-classified frame of the corresponding area; All the boxes to be classified are classified to obtain the bottleneck marker recognition result.

5. The method for intelligent adjustment and image acquisition control of a marker according to claim 4, characterized in that: In the above S2, before identifying the bottleneck marker, it also includes performing image enhancement processing on the bottleneck splicing image to obtain a bottleneck splicing enhanced image for performing the identification operation; The specific operations of image enhancement processing are: If the information entropy of the bottleneck stitching image is greater than the information entropy threshold, the bottleneck stitching image is divided into several sub-regions, and each sub-region is subjected to histogram equalization processing according to the grayscale distribution information of each sub-region until the contrast of each sub-region reaches the corresponding contrast threshold, and then stitching is performed to obtain a bottleneck stitching shallow enhanced image; the pixel value of each position point at the stitching boundary in the bottleneck stitching shallow enhanced image is changed to the average pixel value of the position point within the corresponding four neighborhoods to obtain a bottleneck stitching enhanced image; If the information entropy of the bottleneck stitching image is not greater than the information entropy threshold, the bottleneck stitching image is sequentially processed by closing operation and opening operation to obtain a bottleneck stitching shallow enhanced image; after obtaining the difference map between the bottleneck stitching shallow enhanced image and the bottleneck stitching image, it is superimposed with the bottleneck stitching image to obtain a bottleneck stitching enhanced image.

6. A marker intelligent adjustment and image acquisition control system, characterized in that: The method for realizing the intelligent adjustment and image acquisition control of the marker according to claim 1 comprises: A liftable storage table is used to place the bottled wine to be inspected and drive the bottled wine to be inspected to move; The collection module includes a soft light cover, a collection window, a glass plate, a plane mirror, and a first camera; the height of the soft light cover is a preset wine bottle height, the soft light cover opens downward, the collection window is located on the soft light cover, a plane mirror is obliquely arranged above one side of the collection window, a glass plate is vertically arranged between the collection window and the plane mirror, and a first camera is arranged below the plane mirror; the positional relationship of the collection window, the plane mirror, the glass plate, and the first camera can satisfy the requirement that the first camera can capture the image of the bottled wine to be inspected on the plane mirror passing through the collection window and the glass plate after supplementary light, obtain an initial collection image or a bottleneck marker collection image, and transmit the initial collection image to the processing module; The processing module is used to determine whether there is a complete bottleneck marker in the initial acquisition image after receiving the initial acquisition image from the acquisition module; if so, use the initial acquisition image as the bottleneck marker acquisition image; if not, transmit a signal indicating that there is no complete bottleneck marker in the initial acquisition image to the control module; and after receiving the bottleneck circle image from the second camera, splice the bottleneck circle image to obtain a bottleneck splicing image; based on the position of the bottleneck marker in the bottleneck splicing image, as well as the position of the second camera and the acquisition window, obtain the adjustment direction and adjustment angle of the bottled wine to be inspected, and transmit the signal to the control module; The control module is used to control the up and down movement of the liftable storage platform, and when receiving a signal from the processing module that there is no complete bottleneck marker in the initial acquisition image, control the rotation of the bottled wine to be inspected, so that the second camera obtains a circle image of the bottleneck of the bottled wine to be inspected, and the installation height of the second camera is a preset wine bottle height; and according to the direction and angle of the bottled wine to be inspected transmitted by the processing module, the bottled wine to be inspected is rotated until the bottleneck marker is facing the acquisition window, and then the first camera is controlled to take a picture.

7. The marker intelligent adjustment and image acquisition control system according to claim 6, characterized in that: A first supplementary light source is obliquely arranged on both sides above the soft light cover, and the illumination direction of the first supplementary light source is directly facing the side wall of the soft light cover; a vertical second supplementary light source is arranged between the collection window and the glass plate.

8. The marker intelligent adjustment and image acquisition control system according to claim 6, characterized in that: When the bottled wine to be inspected rises to the preset bottle height, the top height of the bottleneck of the bottled wine to be inspected does not exceed the top height of the collection window, and the bottom height of the bottleneck of the bottled wine to be inspected does not fall below the bottom height of the collection window.

Images