阅读说明：本技术 一种led拼接屏图像的模组划分方法以及led屏校正方法 (Module division method for LED spliced screen image and LED screen correction method ) 是由 王利文 于 2021-09-28 设计创作，主要内容包括：本发明涉及LED屏校正技术领域,具体公开了一种LED拼接屏图像的模组划分方法以及LED屏校正方法,模组划分方法包括获取LED拼接屏的灯点点亮图像；LED拼接屏为若干个模组拼接而成；根据预设起始灯点在X方向的单个图像坐标值以及预设的模组划分策略,在X方向对预设起始灯点所在的一行点亮灯点逐一进行模组划分；根据预设起始灯点在Y方向的单个图像坐标值以及预设的模组划分策略,在Y方向对预设起始灯点所在的一列点亮灯点逐一进行模组划分等步骤。本发明解决了当前LED拼接屏的模组之间修缝效果不佳或修缝错误的问题,进而优化了修缝效果,也进一步优化了LED屏校正效果。(The invention relates to the technical field of LED screen correction, and particularly discloses a module division method of an LED spliced screen image and an LED screen correction method, wherein the module division method comprises the steps of obtaining a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules; according to a single image coordinate value of a preset starting lamp point in the X direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located one by one in the X direction; and according to the single image coordinate value of the preset starting lamp point in the Y direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located in the Y direction one by one, and the like. The invention solves the problem of poor seam repair effect or seam repair error between modules of the current LED spliced screen, thereby optimizing the seam repair effect and further optimizing the correction effect of the LED screen.)

1. A module dividing method for an LED spliced screen image is characterized by comprising the following steps:

acquiring a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules;

according to a single image coordinate value of a preset starting lamp point in the X direction and a preset module division strategy, carrying out module division on a row of lighting lamp points in which the preset starting lamp point is located one by one in the X direction;

according to a single image coordinate value of a preset starting lamp point in the Y direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located one by one in the Y direction;

wherein the module division strategy comprises:

judging whether the current lamp point and the target starting lamp point belong to the same module or not according to the number of the lighting lamp points at a distance from the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the width of the module and the height of the module;

if so,

dividing the current lamp point into the modules where the target starting lamp point is located;

if not, the user can not select the specific application,

dividing the current lamp point into the next module, and setting the current lamp point as the calculation starting lamp point of the module; setting the front lighting point of the calculation starting lamp point in the selected direction as the calculation tail lamp point of the previous module;

the target starting lamp point is the preset starting lamp point or the calculation starting lamp point which is closest to the current lamp point in the forward direction in the selected direction.

2. The method for partitioning the LED tiled screen image into modules according to claim 1, wherein said obtaining the lamp lighting image of the LED tiled screen comprises:

acquiring a plurality of lightening images of the LED spliced screen when the LED spliced screen is partially lightened according to the preset number of jumping points;

and using the plurality of lighting images as the lamp lighting images respectively.

3. The method for dividing LED tiled screen image into modules according to claim 1, wherein said determining whether the current lamp point and the target starting lamp point belong to the same module according to the number of lighting lamp points of the current lamp point and the target starting lamp point, and the preset adjacent lamp point distance, module width and module height comprises:

judging whether x0+ xNum xDelta is less than w; wherein X0 is a single image coordinate value of the target starting lamp point in the X direction, xNum is the number of the lighting lamp points between the current lamp point and the target starting lamp point, xDelta is the distance between two adjacent lighting lamp points in the X direction in the same module, and w is the module width;

and the number of the first and second groups,

judging whether y0+ yNum yDelta is smaller than h; y0 is the single image coordinate value of the target starting lamp point in the Y direction, yNum is the number of the lighting lamp points between the current lamp point and the target starting lamp point, ydal is the distance between two adjacent lighting lamp points in the Y direction in the same module, and h is the module height.

4. The method for partitioning a module of an image of an LED tiled screen according to claim 3, wherein the module partitioning strategy further comprises:

calculating the image coordinates of each starting light point and each tail light point in the X direction according to the single image coordinate value of the preset starting light point in the X direction; and

and calculating the image coordinates of each starting light point and each ending light point in the Y direction according to the single image coordinate value of the preset starting light point in the Y direction.

5. The method for die-set division of an image of an LED tiled screen according to claim 4, wherein calculating the image coordinates of each estimated start light point in the X direction based on the single image coordinate values of the preset start light point in the X direction, and calculating the image coordinates of each estimated start light point in the Y direction based on the single image coordinate values of the preset start light point in the Y direction comprises:

let the image coordinates of the estimated starting lamp point in the X direction be equal to X0+ xNum 1X Delta-w; wherein, xNum1 is the number of lighting lamps of a module in the X direction;

making the image coordinate of the starting lamp point calculated in the Y direction equal to Y0+ yNum1 x yDelta-h; where yNum1 is the number of lighting lamps in the Y direction for one module.

6. The method for partitioning LED tiled screen image into modules according to claim 4, wherein calculating the image coordinates of each estimated end light point in the X direction according to the single image coordinate value of the preset start light point in the X direction, and calculating the image coordinates of each estimated end light point in the Y direction according to the single image coordinate value of the preset start light point in the Y direction comprises:

let the image coordinate of the last estimated light point in the X direction equal to X0+ xNum 2X data delta; wherein xNum2 is the number of lighting lamp points between the estimated tail lamp point and the preset starting lamp point or the estimated starting lamp point in the current module;

making the image coordinate of the estimated tail light point in the Y direction equal to Y0+ yNum2 x yDelta; and the yNum2 is the number of the lighting lamp points which are separated from the estimated tail lamp point and the preset starting lamp point or the estimated starting lamp point in the current module.

7. The method for partitioning LED tiled screen image module according to claim 1,

when the lighting lamp points in the row where the preset starting lamp points are located are subjected to module division one by one in the X direction, an X-direction module array is established and updated, and the expression form of the X-direction module array is as follows: module name [ number of lamps in X direction ];

when the Y direction is right preset initial lamp point place a list of light point and carry out the module and divide one by one, establish and update Y direction module array, the expression form of Y direction module array is: module name [ number of light points in Y direction ].

8. A LED screen correction method based on local distortion elimination is characterized by comprising the following steps:

the module division method according to any one of claims 1 to 7, wherein the module division is performed in the X direction and the Y direction on a lamp lighting image of the LED tiled screen;

acquiring image coordinates of a calculation starting lamp point and a calculation tail lamp point in each module as first image coordinates;

calibrating the camera according to the first image coordinates and the first physical coordinates of the initial light points and the final light points calculated in every two adjacent modules to obtain an internal reference matrix and distortion parameters of the camera;

regarding the two corresponding modules as one module, and acquiring a second image coordinate and a second physical coordinate of the initial light point and the final light point;

carrying out distortion elimination processing on second image coordinates of the initial calculation lamp point and the final calculation lamp point through the camera internal reference matrix and the distortion parameter to generate third image coordinates;

calculating a homography transformation matrix according to the third image coordinate and the second physical coordinate;

and performing homographic transformation on the third image coordinate of the estimated last lamp point of the previous module and the third image coordinate of the estimated starting lamp point of the next module through the homographic transformation matrix to generate a third physical coordinate corresponding to the estimated starting lamp point and the estimated last lamp point.

9. The utility model provides a device is divided to module of LED concatenation screen image which characterized in that, the module divides the device to include that the lamp lights image acquisition module, X direction module and Y direction module and divides the module, wherein:

the lamp lighting image acquisition module is respectively connected with the X-direction module dividing module and the Y-direction module dividing module; the lamp lighting image acquisition module is used for acquiring a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules;

the X-direction module dividing module is connected with the lamp lighting image acquisition module and is used for carrying out module division on a row of lighting lamp points in which a preset starting lamp point is located one by one in the X direction according to a single image coordinate value of the preset starting lamp point in the X direction and a preset module dividing strategy;

the Y-direction module dividing module is connected with the lamp lighting image acquisition module and is used for carrying out module division on a row of lighting lamp points in which the preset starting lamp points are located one by one in the Y direction according to a single image coordinate value of the preset starting lamp points in the Y direction and a preset module dividing strategy;

wherein the module division strategy comprises:

judging whether the current lamp point and the target starting lamp point belong to the same module or not according to the number of the lighting lamp points at a distance from the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the width of the module and the height of the module;

if so,

dividing the current lamp point into the modules where the target starting lamp point is located;

if not, the user can not select the specific application,

dividing the current lamp point into the next module, and setting the current lamp point as the calculation starting lamp point of the module; setting the front lighting point of the calculation starting lamp point in the selected direction as the calculation tail lamp point of the previous module;

the target starting lamp point is the preset starting lamp point or the calculation starting lamp point which is closest to the current lamp point in the forward direction in the selected direction.

10. An LED screen correction system based on local distortion elimination, comprising the module dividing device of claim 9, and a camera, an LED tiled screen, and a correction device, wherein:

the LED spliced screen is used for partially lighting according to the preset jumping point number;

the camera is connected with the module dividing device and used for photographing the LED spliced screen to obtain a plurality of lightening images;

the module dividing device is connected with the camera and the correcting device, and is used for respectively taking a plurality of the lighting images as lamp lighting images and carrying out module division on the lamp lighting images in the X direction and the Y direction;

the correction device is connected with the module dividing device and is used for acquiring the image coordinates of the initial calculation lamp point and the final calculation lamp point in each module as first image coordinates; calibrating the camera according to the first image coordinates and the first physical coordinates of the initial light points and the final light points calculated in every two adjacent modules to obtain an internal reference matrix and distortion parameters of the camera; regarding the two corresponding modules as one module, and acquiring a second image coordinate and a second physical coordinate of the initial light point and the final light point; carrying out distortion elimination processing on second image coordinates of the initial calculation lamp point and the final calculation lamp point through the camera internal reference matrix and the distortion parameter to generate third image coordinates; calculating a homography transformation matrix according to the third image coordinate and the second physical coordinate; and performing homographic transformation on the third image coordinate of the estimated last lamp point of the previous module and the third image coordinate of the estimated starting lamp point of the next module through the homographic transformation matrix to generate a third physical coordinate corresponding to the estimated starting lamp point and the estimated last lamp point.

Technical Field

The invention relates to the technical field of LED screen correction, in particular to a module division method of an LED spliced screen image and an LED screen correction method.

Background

In the prior art, an industrial camera is usually used to photograph an LED display screen to be corrected, so as to complete the correction processing of the LED display screen through correction software, but because the distortion of a picture photographed by the industrial camera is large, especially an industrial camera with a wide-angle lens is serious, the problem that the position of a lamp point extracted from a photometric picture photographed by the industrial camera by the correction software is inaccurate, that is, the problem that the physical coordinate obtained by back-projecting the extracted position of the lamp point does not accord with the actual physical coordinate of the lamp point, and the problem that the seam repair effect between modules in the LED display screen is poor or the seam repair is wrong occurs, so that the correction effect of the LED display screen is directly affected.

Therefore, there is a need to find a new technical solution to solve the above problems.

Disclosure of Invention

Aiming at the technical problems in the prior art, the invention provides a module division method of an LED spliced screen image, an LED screen correction method based on local distortion elimination, a module division device and an LED screen correction system.

The invention discloses a method for dividing modules of an LED spliced screen image, which comprises the following steps:

acquiring a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules;

according to a single image coordinate value of a preset starting lamp point in the X direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located one by one in the X direction;

according to a single image coordinate value of a preset starting lamp point in the Y direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located one by one in the Y direction;

wherein, the module divides the tactics to include:

judging whether the current lamp point and the target starting lamp point belong to the same module or not according to the number of the lighting lamp points at a distance from the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the width of the module and the height of the module;

if so,

dividing the current lamp point into modules where the target starting lamp point is located;

if not, the user can not select the specific application,

dividing the current lamp point into the next module, and setting the current lamp point as the calculation starting lamp point of the module; setting the calculation starting lamp point on the front lighting lamp point in the selected direction as the calculation tail lamp point of the previous module;

the target starting lamp point is a preset starting lamp point or a calculated starting lamp point which is closest to the current lamp point in the forward direction in the selected direction.

Further, the lamp that obtains LED concatenation screen lights up the image, includes:

acquiring a plurality of lightening images of the LED spliced screen when the LED spliced screen is partially lightened according to the preset number of jumping points;

the plurality of lighting images are respectively used as lamp lighting images.

Further, according to the number of the lighting lamp points between the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the module width and the module height, whether the current lamp point and the target starting lamp point belong to the same module or not is judged, including:

judging whether x0+ xNum xDelta is less than w; wherein X0 is a single image coordinate value of the target starting lamp point in the X direction, xNum is the number of the lighting lamp points between the current lamp point and the target starting lamp point, xDelta is the distance between two adjacent lighting lamp points in the X direction in the same module, and w is the module width;

and the number of the first and second groups,

judging whether y0+ yNum yDelta is smaller than h; y0 is the single image coordinate value of the target starting lamp point in the Y direction, yNum is the number of the lighting lamp points between the current lamp point and the target starting lamp point, ydal is the distance between two adjacent lighting lamp points in the Y direction in the same module, and h is the module height.

Further, the module division strategy further includes:

calculating the image coordinates of each starting light point and each tail light point in the X direction according to the single image coordinate value of the preset starting light point in the X direction; and

and calculating the image coordinates of each starting light point and each ending light point in the Y direction according to the single image coordinate value of the preset starting light point in the Y direction.

Further, calculating an image coordinate of each estimated starting light point in the X direction according to a single image coordinate value of the preset starting light point in the X direction, and calculating an image coordinate of each estimated starting light point in the Y direction according to a single image coordinate value of the preset starting light point in the Y direction, includes:

the image coordinates of the starting lamp point calculated in the X direction are equal to X0+ xNum1 xDelta-w; wherein, xNum1 is the number of lighting lamps of a module in the X direction;

making the image coordinate of the starting lamp point calculated in the Y direction equal to Y0+ yNum1 x yDelta-h; where yNum1 is the number of lighting lamps in the Y direction for one module.

Further, calculating an image coordinate of each estimated last light point in the X direction according to a single image coordinate value of the preset start light point in the X direction, and calculating an image coordinate of each estimated last light point in the Y direction according to a single image coordinate value of the preset start light point in the Y direction, including:

making the image coordinate of the tail lamp point calculated in the X direction equal to X0+ xNum 2X Delta; wherein, xNum2 is used for calculating the number of the lighting lamp points between the tail lamp points and the preset starting lamp points in the current module or calculating the distance between the starting lamp points;

making the image coordinate of the tail light point calculated in the Y direction equal to Y0+ yNum2 x yDelta; wherein, yNum2 is the number of lighting lamp points for calculating the distance between the tail lamp point and the preset starting lamp point in the current module or calculating the distance between the starting lamp points.

Further, the method also comprises the following steps:

when the X direction carries out the module to predetermineeing the one line light point that initial lamp point place and divide one by one, establish and update X direction module array, the expression form of X direction module array is: module name [ number of lamps in X direction ];

when carrying out the module partition one by one to presetting a row of lighting lamp point at originated lamp point in Y direction, establish and update Y direction module array, the expression form of Y direction module array is: module name [ number of light points in Y direction ].

The invention also comprises an LED screen correction method based on local distortion elimination, which comprises the following steps:

dividing the modules in the X direction and the Y direction of a lamp lighting image of the LED spliced screen according to the module dividing method;

acquiring image coordinates of a calculation starting lamp point and a calculation tail lamp point in each module as first image coordinates;

calibrating the camera according to the first image coordinates and the first physical coordinates of the initial light points and the final light points calculated in every two adjacent modules to obtain an internal reference matrix and distortion parameters of the camera;

regarding the two corresponding modules as one module, and acquiring a second image coordinate and a second physical coordinate of the initial light point and the final light point;

carrying out distortion elimination processing on second image coordinates of the initial calculation lamp point and the final calculation lamp point through the camera internal reference matrix and the distortion parameters to generate third image coordinates;

calculating a homography transformation matrix according to the third image coordinate and the second physical coordinate;

and performing homographic transformation on the third image coordinate of the estimated last lamp point of the previous module and the third image coordinate of the estimated starting lamp point of the next module through a homographic transformation matrix to generate a third physical coordinate corresponding to the estimated starting lamp point and the estimated last lamp point.

The invention also comprises a module dividing device for the LED spliced screen image, which comprises a lamp lighting image acquisition module, an X-direction module dividing module and a Y-direction module dividing module, wherein the lamp lighting image acquisition module comprises:

the lamp lighting image acquisition module is respectively connected with the X-direction module dividing module and the Y-direction module dividing module; the lamp lighting image acquisition module is used for acquiring a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules;

the X-direction module dividing module is connected with the lamp lighting image acquisition module and used for carrying out module division on a row of lighting lamp points where the preset starting lamp points are located one by one in the X direction according to a single image coordinate value of the preset starting lamp points in the X direction and a preset module dividing strategy;

the Y-direction module dividing module is connected with the lamp lighting image acquisition module and is used for carrying out module division on a row of lighting lamp points in which the preset starting lamp points are located one by one in the Y direction according to a single image coordinate value of the preset starting lamp points in the Y direction and a preset module dividing strategy;

wherein, the module divides the tactics to include:

judging whether the current lamp point and the target starting lamp point belong to the same module or not according to the number of the lighting lamp points at a distance from the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the width of the module and the height of the module;

if so,

dividing the current lamp point into modules where the target starting lamp point is located;

if not, the user can not select the specific application,

dividing the current lamp point into the next module, and setting the current lamp point as the calculation starting lamp point of the module; setting the calculation starting lamp point on the front lighting lamp point in the selected direction as the calculation tail lamp point of the previous module;

the target starting lamp point is a preset starting lamp point or a calculated starting lamp point which is closest to the current lamp point in the forward direction in the selected direction.

The invention also comprises an LED screen correction system based on local distortion elimination, wherein the LED screen correction system comprises the module dividing device, a camera, an LED spliced screen and a correction device, wherein:

the LED spliced screen is used for partially lighting according to the preset jumping point number;

the camera is connected with the module dividing device and used for photographing the LED spliced screen to obtain a plurality of lightening images;

a module dividing device connected to the camera and the correcting device, the module dividing device being configured to take the plurality of lighting images as lamp lighting images, and to divide the lamp lighting images into modules in the X direction and the Y direction;

the correction device is connected with the module dividing device and used for acquiring the image coordinates of the initial calculation lamp point and the final calculation lamp point in each module as first image coordinates; calibrating the camera according to the first image coordinates and the first physical coordinates of the initial light points and the final light points calculated in every two adjacent modules to obtain an internal reference matrix and distortion parameters of the camera; regarding the two corresponding modules as one module, and acquiring a second image coordinate and a second physical coordinate of the initial light point and the final light point; carrying out distortion elimination processing on second image coordinates of the initial calculation lamp point and the final calculation lamp point through the camera internal reference matrix and the distortion parameters to generate third image coordinates; calculating a homography transformation matrix according to the third image coordinate and the second physical coordinate; and performing homographic transformation on the third image coordinate of the estimated last lamp point of the previous module and the third image coordinate of the estimated starting lamp point of the next module through a homographic transformation matrix to generate a third physical coordinate corresponding to the estimated starting lamp point and the estimated last lamp point.

According to the LED spliced screen image module dividing method, the LED screen correction method based on local distortion elimination, the module dividing device and the LED screen correction system, the lamp points on the full-screen lamp lighting image of the LED spliced screen are divided into the modules from the X direction and the Y direction, so that the starting lamp point and the tail lamp point of each module on the full-screen lamp lighting image are calculated, the distortion elimination of the gaps among the modules in the full-screen lamp lighting image is facilitated, the lamp point image coordinates on two sides of the gaps of the modules are adjusted, the subsequent seam repair processing is performed through the image coordinates of the lamp points, the problem that the seam repair effect between the modules of the current LED spliced screen is poor or wrong is solved, the seam repair effect is optimized, and the LED screen correction effect is further optimized.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a flowchart (I) illustrating a method for dividing an image into modules of an LED tiled screen according to an embodiment of the present invention;

FIG. 2 is a flowchart (II) illustrating steps of a method for dividing an image module of an LED tiled screen according to an embodiment of the present invention;

FIG. 3 is a flow chart (III) illustrating steps of a method for dividing an image module of an LED tiled screen according to an embodiment of the present invention;

FIG. 4 is a flowchart (IV) illustrating steps of a module division method for an image of an LED tiled screen according to an embodiment of the present invention;

FIG. 5 is a flowchart illustrating steps of a method for correcting an LED screen based on local distortion elimination according to an embodiment of the present invention;

FIG. 6 is a structural composition diagram of a module dividing device for an LED tiled screen image according to an embodiment of the present invention;

fig. 7 is a structural composition diagram of an LED screen correction system based on local distortion elimination according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In order to meet the image display requirements of general crowds, the conventional large and medium LED display screen is formed by splicing a plurality of LED modules of the same type, the lamp points in each LED module are uniformly arranged at equal intervals, and the interval between the modules is too narrow or too wide compared with the interval between the lamp points in the modules due to the influence of splicing. When the LED display screen is corrected, if the camera calibration is directly carried out by taking the lamp dot matrix of the whole LED screen as the checkerboard, the distortion elimination treatment is realized, the distortion elimination effect and the actual condition still have deviation because the distances among the modules are unequal, at the moment, each module can be taken as the checkerboard under a specific view field by utilizing the equal distances among the lamp dots in the module, the camera calibration is carried out on the checkerboard, the distortion elimination treatment is carried out, however, the module division of the lamp dots in the LED spliced screen image needs to be completed firstly, and the lamp dots are also distributed into the corresponding modules.

Therefore, an embodiment of the present invention first provides a method for dividing a module of an image of an LED tiled screen, as shown in fig. 1, including:

step S10: acquiring a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules.

The method is suitable for module division of the image of the spliced LED display screen, and the lamp lighting image of the LED spliced screen obtained in the step is an image of a lamp with a lighting state.

Specifically, as shown in fig. 3, in step S10, the acquiring a lamp lighting image of the LED tiled screen includes:

step S101: and acquiring a plurality of lightening images of the LED spliced screen when the LED spliced screen is partially lightened according to the preset number of jumping points.

In order to ensure the correction effect of the subsequent scheme on the LED screen, the step needs to acquire a plurality of images of partial lamps which are lighted. The parameters related to this step include a preset "number of hops", for example, the value of "number of hops" is equal to 2, which can indicate that when the current lamp is lit, the lamp that is 2 lamp points away from the current lamp is simultaneously lit, i.e., the two lamp points are separated by 1 unlit lamp point. The specific value of the "number of hops" can be set by a person skilled in the art when the method is implemented, and it is preferable that the lighted lamp points in the taken image do not interfere with each other.

Step S102: the plurality of lighting images are respectively used as lamp lighting images.

When all the lamp points are lit in different lighting images, each lighting image is taken as the lamp lighting image of the present embodiment, and the subsequent processing is performed separately.

After the lamp lighting image of the LED tiled screen is obtained in step S10, the lighting lamp points on the lamp lighting image are divided into modules, and step S20 and step S30 are respectively divided into modules in the X direction and the Y direction, and the execution sequence of these two steps is not limited, that is, the module division in the X direction is performed first, then the module division in the Y direction is performed, and then the module division in the Y direction is performed first, then the module division in the X direction is performed, which may be the first mode, which is only used as an example for the present invention.

Step S20: and according to the single image coordinate value of the preset starting lamp point in the X direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located one by one in the X direction.

Step S30: and according to the single image coordinate value of the preset starting lamp point in the Y direction and a preset module division strategy, carrying out module division on a row of lighting lamp points where the preset starting lamp point is located one by one in the Y direction.

The module dividing method provided by the embodiment of the invention comprises the dividing processes in the X direction and the Y direction, the specific dividing modes are basically the same, and the difference is only in the direction in which the lighting lamp points are divided one by one. As shown in fig. 2, the module division policy includes:

step S01: and judging whether the current lamp point and the target starting lamp point belong to the same module or not according to the number of the lighting lamp points at a distance from the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the width of the module and the height of the module.

If yes, go to step S02:

step S02: and dividing the current lamp point into modules where the target starting lamp point is located.

In this step, the number of the lighting lamps of the module in the selected direction is increased by one, so as to update the number of the lighting lamps of the module.

If not, go to step S03:

step S03: dividing the current lamp point into the next module, and setting the current lamp point as the calculation starting lamp point of the module; and setting the estimation starting lamp point to be a previous lighting point in the selected direction as the estimation ending lamp point of the previous module.

The target starting lamp point in the module division strategy is a preset starting lamp point or a calculated starting lamp point which is closest to the current lamp point in the forward direction in the selected direction.

Generally, the first lighting lamp of the first module is used as a default starting lamp having known image coordinates (x and y values), and the known values in this step further include default adjacent lamp spacing, module width, and module height.

The lighting lamp points in a row (column) are divided one by one in the X direction (or Y direction), first by the first lighting lamp point after (below) the preset starting lamp point, and second, third, and fourth … …. When the current lamp point and the target starting lamp point belong to the same module, dividing the current lamp point into the modules where the target starting lamp point is located, and adding one to the number of the lighting lamp points of the module in the selected direction; if the current lamp point and the target starting lamp point are different and belong to one module, the current lamp point is divided into the next module, the current lamp point is set as the calculation starting lamp point of the module to which the current lamp point belongs, the previous lamp point of the calculation starting lamp point in the selected direction is judged to belong to the previous module before the lamp point, and therefore the previous lamp point of the calculation starting lamp point in the selected direction is set as the calculation tail lamp point of the previous module. By this method, each module and the estimated start lamp point and the estimated end lamp point of each module are divided in the selected direction.

Specifically, in the embodiment of the present invention, step S01: according to the number of the lighting lamp points between the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the module width and the module height, whether the current lamp point and the target starting lamp point belong to the same module or not is judged, and the method comprises the following steps:

step S011: judging whether x0+ xNum xDelta is less than w; wherein, X0 is the initial lamp point of target single image coordinate value in the X direction, and xNUM is the light point quantity of lighting that the initial lamp point of current lamp point and target was apart from, and xDelta is the interval of two adjacent light points of lighting in the X direction in the same module, and w is the module width. And the number of the first and second groups,

step S012: judging whether y0+ yNum yDelta is smaller than h; y0 is the single image coordinate value of the target starting lamp point in the Y direction, yNum is the number of the lighting lamp points between the current lamp point and the target starting lamp point, ydal is the distance between two adjacent lighting lamp points in the Y direction in the same module, and h is the module height.

The values of xDelta and yDelta are obtained by multiplying the distance between two adjacent lamp points and the number of the jumping points. X0 is a single image coordinate value of the target start lamp point in the X direction, that is, a preset start lamp point closest to the current lamp point in the forward direction or a single image coordinate value of the start lamp point in the X direction calculated, for example, the number of the lighting lamp points in the LED tiled screen is 8 × 8, and it is composed of 4 modules, that is, 2 × 2 modules, that is, two modules in the X direction, and two modules in the Y direction. Setting the serial numbers of the lighted lamp points in the X direction to 0,1, 2, 3, 4, 5, 6, and 7, respectively, so that the lamp points 0,1, 2, and 3 belong to the first module, and 4, 5, 6, and 7 belong to the second module, in the lamp lighting image, setting the lamp point 0 as a preset starting lamp point, and the coordinates thereof are set to (X0, y 0), so that the number of the lighted lamp points between the first lighted lamp point after the preset starting lamp point and the preset starting lamp point is 1, that is, there is a lamp point distance of one unit between the two lighted lamp points 0 and 1, calculating the relationship between X0+1 xDelta and the module width w, and certainly determining that X0+1 xDelta is smaller than w, for the following reasons: given that the number of lit spots per module is 4X 4, then w =3 units of spot spacing + first spot to module left edge distance + last spot to module right edge distance, either in the X direction or the Y direction. Thus, the lighting lamp points are judged one by one, and the lighting lamp points belong to the next module only when the current lamp points meet the condition that x0+ xNum xDelta is greater than or equal to w. The principle of step S012 is the same as that of step S011, and is not described here again.

Specifically, in the embodiment of the present invention, on the basis of the above embodiment, the module division policy further includes:

step S04: and calculating the image coordinates of each starting light point and each ending light point in the X direction according to the single image coordinate value of the preset starting light point in the X direction.

Step S05: and calculating the image coordinates of each starting light point and each ending light point in the Y direction according to the single image coordinate value of the preset starting light point in the Y direction.

By the embodiment of the method, the lighting lamp points in the lamp lighting image are divided into modules, and the image coordinates of each of the estimated initial lamp point and the estimated final lamp point are calculated.

Specifically, in step S04 of the embodiment of the present invention: calculating image coordinates of each estimated start light point in the X direction based on a single image coordinate value of the preset start light point in the X direction, and in step S05: calculating the image coordinate of each calculation starting lamp point in the Y direction according to the single image coordinate value of the preset starting lamp point in the Y direction, wherein the calculation starting lamp point calculation method respectively comprises the following steps:

the image coordinates of the starting lamp point calculated in the X direction are equal to X0+ xNum1 xDelta-w; wherein, xNum1 is the number of lighting lamp points of a module in the X direction.

In this step, the newly determined estimated start lamp point is regarded as the current lamp point, and X0 is still defined as the single image coordinate value of the target start lamp point in the X direction, where the target start lamp point is defined as the preset start lamp point or the estimated start lamp point closest to the current lamp point in the forward direction in the selected direction, as in the previous embodiment. Therefore, for the newly determined estimated start lamp point, the image coordinates in the X direction are: the sum of the X0 and xNum 1X × date X of the single image coordinate value of the preceding estimated starting lamp point (which may be the preset starting lamp point) in the X direction, and the width w of the module is subtracted, and the determined image coordinate of the estimated starting lamp point in the X direction becomes X0 when the image coordinate of the next newly determined estimated starting lamp point in the X direction is calculated.

Making the image coordinate of the starting lamp point calculated in the Y direction equal to Y0+ yNum1 x yDelta-h; where yNum1 is the number of lighting lamps in the Y direction for one module.

The manner of calculating the image coordinates of the starting light point calculated in the Y direction is the same as the principle of calculating the image coordinates of the starting light point calculated in the X direction, and is not described herein again.

Specifically, in step S04 of the embodiment of the present invention: calculating image coordinates of each estimated end light point in the X direction based on a single image coordinate value of the preset start light point in the X direction, and in step S05: calculating the image coordinate of each estimated tail light point in the Y direction according to the single image coordinate value of the preset starting light point in the Y direction, wherein the image coordinate comprises the following steps:

making the image coordinate of the tail lamp point calculated in the X direction equal to X0+ xNum 2X Delta; wherein, xNum2 is that calculating the number of lighting lamp points that the tail lamp point presets the initial lamp point or calculates the initial lamp point apart from in the current module.

Assuming that the light point 0 is the preset initial light point in the current module, the light point 3 in the module is determined as the estimated end light point through the above calculation process, so the image coordinate of the light point 3 is x0+3 x digital data.

Making the image coordinate of the tail light point calculated in the Y direction equal to Y0+ yNum2 x yDelta; wherein, yNum2 is the number of lighting lamp points for calculating the distance between the tail lamp point and the preset starting lamp point in the current module or calculating the distance between the starting lamp points.

The principle of calculating the image coordinates of the tail light points calculated in the Y direction is the same as that of calculating the image coordinates of the tail light points calculated in the X direction, and the description thereof is omitted.

Specifically, on the basis of the above embodiments, as shown in fig. 4, the module dividing method in the embodiment of the present invention further includes:

step S40: establishing and updating an X-direction module array, wherein the expression form of the X-direction module array is as follows: module name [ number of lamps in X direction ];

step S50: the Y-direction module array is established and updated, and the expression form of the Y-direction module array is as follows: module name [ number of light points in Y direction ].

In the process of dividing the modules by judging the lighting lamp points in the selected direction one by one, if the current lamp point and the target starting lamp point belong to the same module, updating the [ number of lamp points in the X direction ] or the [ number of lamp points in the Y direction ] of the module to which the target starting lamp point belongs, if the current lamp point does not belong to the same module as the target starting lamp point, newly building a module, and expressing that: the second module [1] indicates that the module currently has only one lighting lamp point (i.e., the newly determined starting lamp point), or can be represented as the first module [0] by combining the computer language numbering mode (starting from 0), and is represented as the first module [1] when there are two determined lighting lamp points.

The embodiment of the invention also comprises an LED screen correction method based on local distortion elimination, as shown in FIG. 5, comprising the following steps:

step S100: according to the module division method of the embodiment, the module division in the X direction and the Y direction is carried out on the lamp lighting image of the LED spliced screen.

Step S200: and acquiring the image coordinates of the initial calculation lamp point and the final calculation lamp point in each module as first image coordinates.

In this step, the image coordinates of the estimated start light point and the estimated end light point are determined, and the image coordinates of each of the estimated start light point and the estimated end light point in the X direction and the Y direction can be calculated and determined according to the image coordinates of the preset start light point in steps S04 and S05 of the above embodiment. Those skilled in the art can also determine the coordinates by other coordinate calculation methods, such as identifying the positions of all the lighted lamp points in the lamp lighting image to obtain the pixel coordinates of each lamp point, calculating the module width w 'in the lamp lighting image according to the size of the LED screen and the size of the lamp lighting image, and performing a difference operation between the lamp point pixel coordinates and the module width w' to obtain the image coordinates of each estimated starting lamp point and each estimated ending lamp point.

Step S300: and calibrating the camera according to the first image coordinate and the first physical coordinate of the starting light point and the ending light point calculated in each two adjacent modules to obtain the camera internal reference matrix and the distortion parameter.

Taking an LED screen composed of 2 x 2 modules as an example, two groups of modules adjacent to each other left and right and two groups of modules adjacent to each other up and down exist, each module is taken as a checkerboard under a specific visual field in a camera calibration mode, the first image coordinates of the initial lamp point and the final lamp point are calculated to be known, the first physical coordinates of the initial lamp point and the final lamp point are calculated to be the coordinates on the LED screen, and the coordinates of the lamp points on the LED screen are as follows: for example, the number of the first row of the lamp points is w, the first physical coordinates of the lamp points are (0, 0), (1, 0), (2, 0), (3, 0) … … (w-1, 0), the first physical coordinates of the second row of the lamp points are (0, 1), (1, 1), (2, 1), (3, 1) … … (w-1, 1), the first physical coordinates of the last row of the lamp points are (0, h-1), (1, h-1), (2, h-1), (3, h-1) … … (w-1, h-1), and the total number of the last row of the lamp points is h.

The main purpose of the embodiment of the invention is to eliminate distortion of the gap between the modules in the lamp lighting image, and the gap is repaired based on the distance between the lamp points at the two sides of the gap (the width of the gap), so that the gap is only repaired with respect to the two adjacent modules. And (3) taking each module as a checkerboard under a specific view field (or taking the whole light point matrix as a checkerboard), and calibrating the camera so as to obtain a camera internal parameter matrix A and a distortion parameter D. It should be noted that, in an actual scene, perfect splicing between modules cannot be achieved in a parallel and coplanar manner, and a global camera internal parameter matrix and distortion parameters are used for distortion elimination of lamp points, so that the estimation of the seam width is affected by modules which are not adjacent to the seam, and the seam repair effect is not improved or is not improved a lot, and sometimes even is poor. Therefore, the present invention adopts the present embodiment to perform a local seam width estimation method, i.e., the lamp points of the two modules adjacent to the seam are used as camera calibration, and after distortion elimination processing is performed on the lamp points, the estimation of the seam width is completed.

Step S400: and regarding the two corresponding modules as one module, and acquiring a second image coordinate and a second physical coordinate of the initial light point and the final light point.

When two modules are regarded as one module, the second image coordinate and the second physical coordinate of the former module are the same as the first image coordinate and the first physical coordinate, and the second image coordinate and the second physical coordinate of the latter module are changed compared with the first image coordinate and the first physical coordinate.

Step S500: and carrying out distortion elimination processing on the second image coordinates of the initial light point and the final light point to generate a third image coordinate.

The distortion removal processing of this embodiment may use an existing calibration method, and this embodiment is not limited specifically.

Step S600: and calculating a homography transformation matrix according to the third image coordinate and the second physical coordinate.

Homography transformation (homography transformation) is the mapping relation from one plane to another, and the step is carried out on the image coordinates and the physical coordinates but is carried out with strain transformation to obtain a homography transformation matrix.

Step S700: and performing homographic transformation on the third image coordinate of the estimated last lamp point of the previous module and the third image coordinate of the estimated starting lamp point of the next module through a homographic transformation matrix to generate a third physical coordinate corresponding to the estimated starting lamp point and the estimated last lamp point.

The distance between the modules can be calculated through the third physical coordinate, so that the brightness of the LED screen required to be adjusted can be determined through the distance, and then the LED screen is corrected according to the brightness. The subsequent LED screen calibration method can be implemented with reference to the prior art, and is not limited herein.

The embodiment of the present invention further provides a module dividing device 10 for an image of an LED tiled screen, as shown in fig. 6, the module dividing device 10 includes a lamp lighting image obtaining module 101, an X-direction module dividing module 102, and a Y-direction module dividing module 103, where:

a lamp lighting image acquisition module 101 connected to the X-direction module dividing module 102 and the Y-direction module dividing module 103, respectively; the lamp lighting image acquisition module 101 is used for acquiring a lamp lighting image of the LED spliced screen; the LED spliced screen is formed by splicing a plurality of modules;

an X-direction module dividing module 102 connected to the lamp lighting image obtaining module 101, wherein the X-direction module dividing module 102 is configured to perform module division on a row of lighting lamp points in which a preset starting lamp point is located one by one in the X direction according to a single image coordinate value of the preset starting lamp point in the X direction and a preset module dividing policy;

a Y-direction module dividing module 103 connected to the lamp lighting image obtaining module 101, wherein the Y-direction module dividing module 103 is configured to perform module division on a row of lighting lamp points in which a preset starting lamp point is located one by one in the Y direction according to a single image coordinate value of the preset starting lamp point in the Y direction and a preset module dividing policy;

wherein, the module divides the tactics to include:

judging whether the current lamp point and the target starting lamp point belong to the same module or not according to the number of the lighting lamp points at a distance from the current lamp point and the target starting lamp point, and the preset distance between the adjacent lamp points, the width of the module and the height of the module;

if so,

dividing the current lamp point into modules where the target starting lamp point is located;

if not, the user can not select the specific application,

dividing the current lamp point into the next module, and setting the current lamp point as the calculation starting lamp point of the module; setting the calculation starting lamp point on the front lighting lamp point in the selected direction as the calculation tail lamp point of the previous module;

the target starting lamp point is a preset starting lamp point or a calculated starting lamp point which is closest to the current lamp point in the forward direction in the selected direction.

The module dividing device according to the embodiment of the present invention can be implemented by performing the steps of the module dividing method according to the embodiment of the present invention, which will not be described herein again.

The present invention further includes an embodiment of an LED screen correction system based on local distortion removal, as shown in fig. 7, the LED screen correction system includes the module dividing device 10 of the above embodiment, a camera 20, an LED tiled screen 30, and a correction device 40, wherein:

the LED spliced screen 30 is used for partially lighting according to the preset jumping point number;

the camera 20 is connected with the module dividing device 10, and the camera 20 is used for photographing the LED spliced screen 30 to obtain a plurality of lightening images;

the module dividing device 10 is connected to the camera 20 and the correcting device 40, and the module dividing device 10 is configured to use a plurality of lighting images as lamp lighting images, and perform module division in the X direction and the Y direction on the lamp lighting images.

A correcting device 40 connected to the module dividing device 10, wherein the correcting device 40 is configured to obtain image coordinates of the initial estimation light point and the final estimation light point in each module as first image coordinates; calibrating the camera according to the first image coordinates and the first physical coordinates of the initial light points and the final light points calculated in every two adjacent modules to obtain an internal reference matrix and distortion parameters of the camera; regarding the two corresponding modules as one module, and acquiring a second image coordinate and a second physical coordinate of the initial light point and the final light point; carrying out distortion elimination processing on second image coordinates of the initial calculation lamp point and the final calculation lamp point through the camera internal reference matrix and the distortion parameters to generate third image coordinates; calculating a homography transformation matrix according to the third image coordinate and the second physical coordinate; and performing homographic transformation on the third image coordinate of the estimated last lamp point of the previous module and the third image coordinate of the estimated starting lamp point of the next module through a homographic transformation matrix to generate a third physical coordinate corresponding to the estimated starting lamp point and the estimated last lamp point.

The functions implemented by the calibration apparatus 40 according to the embodiment of the present invention can be implemented according to the embodiments of the calibration method described above, and are not described herein again. The module dividing device 10 and the correcting device 40 according to the embodiment of the present invention can be implemented by a computer having corresponding functions, and preferably, the LED tiled screen 30 is connected to the computer in a communication manner, and the number of "hops" is set by the computer, so that the LED tiled screen 30 lights corresponding lamp points.

According to the LED spliced screen image module dividing method, the LED screen correction method based on local distortion elimination, the module dividing device and the LED screen correction system, the lamp points on the LED spliced screen full-screen lamp lighting image are divided into the modules from the X direction and the Y direction, so that the calculation starting lamp point and the calculation tail lamp point in each module on the full-screen lamp lighting image are obtained, the distortion elimination of the gaps among the modules in the full-screen lamp lighting image is facilitated, the lamp point image coordinates on two sides of the gaps of the modules are adjusted, the subsequent seam repair processing is carried out through the image coordinates of the lamp points, the problem that the seam repair effect between the modules of the current LED spliced screen is poor or wrong is solved, the seam repair effect is optimized, and the LED screen correction effect is further optimized.

The present invention has been further described with reference to specific embodiments, but it should be understood that the detailed description should not be construed as limiting the spirit and scope of the present invention, and various modifications made to the above-described embodiments by those of ordinary skill in the art after reading this specification are within the scope of the present invention.