A method for controlling the projection quality of a vehicle-mounted projection device through image correction

Abstract

The present invention discloses a method for image correction control applied to the projection quality of vehicle-mounted projection equipment, which relates to the field of projection control, including: obtaining a normal imaging image of a sample grid image in a projection screen during normal projection; obtaining an actual imaging image; coordinate modeling of the projection screen; obtaining at least one curved edge grid; performing coordinate linearization processing on the curved edge grid, the actual imaging image is corrected to a first corrected image, and at least one deformed grid is obtained; a deformed area is obtained, and the deformed grid is classified in two ways; the area other than the deformed area in the first corrected image is used as a non-deformed area; scaling the deformed area, and translating the non-deformed area. By performing linearization processing and scaling processing, when a demonstration system composed of vehicle-mounted projection equipment encounters geometric distortion, it can perform autonomous identification, and adjust through an algorithm, which can save time and improve the adjustment speed.

Description

Technical Field

[0001] The present invention relates to the field of projection control, and in particular to a method for controlling the image correction of the projection quality of a vehicle-mounted projection device. Background Art

[0002] Adaptive geometric correction of the image of the vehicle-mounted projection device is to automatically correct the image of the vehicle-mounted projection device according to the surface information of a specific type. With the promotion of visualization technology, virtual reality and ubiquitous computing environment, various projection devices with projection optics as the core technology play an important role in daily life. However, when the demonstration system composed of vehicle-mounted projection equipment encounters geometric distortion, it often requires manual adjustment, which is time-consuming and labor-intensive. Summary of the invention

[0003] In order to solve the above technical problems, a method for image correction control applied to the projection quality of vehicle-mounted projection equipment is provided. This technical solution solves the problem that when the demonstration system composed of the vehicle-mounted projection equipment proposed in the above background technology encounters geometric distortion, it often requires manual adjustment, which is time-consuming and labor-intensive.

[0004] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0005] A method for controlling the image correction of projection quality of a vehicle-mounted projection device, comprising:

[0006] Forming a sample grid image for correction testing, and obtaining a normal imaging picture of the sample grid image on the projection screen during normal projection;

[0007] During actual projection, the vehicle-mounted projection device uses the sample grid image for projection, captures the projection image, and obtains the actual imaging image;

[0008] The projection screen is modeled with coordinates, with the lower left corner of the projection screen as the origin, the horizontal direction as the horizontal axis, and the vertical direction as the vertical axis, to obtain the coordinates of the pixel points in the actual imaging picture and the normal imaging picture;

[0009] Recording a grid in the actual imaging picture as a first grid, recording a grid in the normal imaging picture as a second grid, establishing a corresponding relationship between the grids in the actual imaging picture and the grids in the normal imaging picture, comparing the actual imaging picture with the normal imaging picture, and obtaining at least one curved edge grid, where the curved edge grid is the grid in the actual imaging picture;

[0010] The curved edge mesh is subjected to coordinate linearization processing, the actual imaging picture is corrected into a first corrected image, and the first corrected image is compared with the normal imaging picture to obtain at least one deformed mesh;

[0011] Summarize the deformed meshes to obtain the deformed regions, and classify the deformed meshes in two ways: the deformed meshes with the same abscissa at the center of the deformed regions are classified into the first deformed mesh classification, and the deformed meshes with the same ordinate at the center of the deformed regions are classified into the second deformed mesh classification;

[0012] The area other than the deformed area in the first corrected image is regarded as the non-deformed area;

[0013] The deformed area is scaled and the non-deformed area is translated to complete the image correction.

[0014] Preferably, the establishing of a correspondence between the grids in the actual imaging picture and the normal imaging picture comprises the following steps:

[0015] In the actual imaging picture, the number of first grids on the left side of the first grid is counted as the first horizontal number, and the number of first grids on the lower side of the first grid is counted as the first vertical number;

[0016] In a normal imaging picture, the number of second grids on the left side of the second grid is counted as the second horizontal number, and the number of second grids on the lower side of the second grid is counted as the second vertical number;

[0017] A corresponding relationship is established between a first grid and a second grid in which the first horizontal number is equal to the second horizontal number and the first vertical number is equal to the second vertical number.

[0018] Preferably, the step of comparing the actual imaging picture with the normal imaging picture to obtain at least one curved edge mesh comprises the following steps:

[0019] Take at least one first sampling point at the edge of the first grid, perform fitting on the first sampling point, and obtain a first fitting function, where the first fitting function is a function expression of the edge of the first grid;

[0020] Take at least one second sampling point at the edge of the second grid, perform fitting on the second sampling point, and obtain a second fitting function, where the second fitting function is a function expression of the edge of the second grid;

[0021] Using the first curvature integral formula, calculate the first integrated curvature of the first grid;

[0022] Using the second curvature integral formula, calculate the second integrated curvature of the second grid;

[0023] When the difference between the first comprehensive curvature and the second comprehensive curvature of the corresponding first grid and second grid is greater than a preset value, the first grid is used as a curved edge grid;

[0024] The first curvature integral formula is as follows: ,

[0025] Among them, the integral path is the path corresponding to the edge of the first grid, A is the first comprehensive curvature, is the second-order derivative of the first fitting function, is the first derivative of the first fitting function, and x is the abscissa of a point on the edge of the first grid;

[0026] The second curvature integral formula is as follows: ,

[0027] Among them, the integral path is the path corresponding to the edge of the second grid, B is the second comprehensive curvature, is the second derivative of the second fitting function, is the first derivative of the second fitting function, and t is the abscissa of the point on the edge of the second grid.

[0028] Preferably, the coordinate linearization processing is performed on the curved edge grid, and the actual imaging picture is corrected to the first corrected image, which includes the following steps:

[0029] Obtain the first grid whose center coordinates have the smallest difference with the center of the curved edge grid as the first characteristic grid;

[0030] Obtaining the difference between the ordinate of a point on the upper edge of the first feature grid and the ordinate of the center of the first feature grid as the first distance;

[0031] Obtain the difference between the abscissa of a point on the left edge of the first feature grid and the abscissa of the center of the first feature grid as the second distance;

[0032] The coordinates of the points in the curved edge grid are linearized using a linearization formula, and the actual imaging image is corrected to a first corrected image;

[0033] The linearization formula is as follows: ,

[0034] Among them, m and n are the horizontal and vertical coordinates of the center of the curved edge grid, k and l are the horizontal and vertical coordinates of the points in the curved edge grid, respectively. is the first fitting function, for is the inverse function of the linearized curved edge mesh, where D is the first distance, d is the second distance, and K and L are the horizontal and vertical coordinates of the points in the linearized curved edge mesh, respectively.

[0035] Preferably, the step of comparing the first corrected image with the normal imaging picture to obtain at least one deformed mesh comprises the following steps:

[0036] The grid in the first corrected image is used as the third grid, and the grid correspondence relationship of the first corrected image before correction is used to correspond the third grid to the second grid;

[0037] The area of ​​the third grid is compared with the area of ​​the corresponding second grid. When the difference between the area of ​​the third grid and the area of ​​the corresponding second grid is greater than a preset value, the third grid is used as a deformable grid.

[0038] Preferably, the scaling of the deformation area comprises the following steps:

[0039] Forming a ordinate interval for the first deformed mesh classification, wherein the ordinate interval is a value range of the ordinates of the deformed meshes in the first deformed mesh classification;

[0040] The ratio of the side length of the upper side of the deformed mesh in the first deformed mesh classification to the side length of the upper side of the corresponding second mesh is used as the first coefficient;

[0041] forming a horizontal coordinate interval for the second deformed mesh classification, wherein the horizontal coordinate interval is a value range of the horizontal coordinates of the deformed meshes in the second deformed mesh classification;

[0042] The ratio of the side length of the left side of the deformed mesh in the second deformed mesh classification to the side length of the left side of the corresponding second mesh is used as the second coefficient;

[0043] Randomly select a point in the deformation area as the deformation point, obtain the first deformed mesh classification located on the left side of the deformation point as the first target classification, and obtain the first deformed mesh classification containing the deformation point as the first feature classification;

[0044] Acquire a second deformed mesh classification located below the deformation point as a second target classification, and acquire a second deformed mesh classification including the deformation point as a second feature classification;

[0045] Use the scaling formula to calculate the scaled coordinates of the deformation point. When scaling, scale the deformation point to the scaled coordinates of the deformation point.

[0046] The scaling formula is as follows: ,

[0047] Among them, O is the ordinate of the scaled coordinate of the deformation point, o is the ordinate of the deformation point, is the first coefficient of the first feature classification, s is the left endpoint of the ordinate interval of the first feature classification, i is the counting index, r is the number of the first target classification, is the length of the ordinate interval of the i-th first target classification, is the first coefficient of the i-th first target classification, P is the abscissa of the scaled coordinates of the deformation point, and p is the abscissa of the deformation point. is the second coefficient of the second feature classification, u is the left endpoint of the horizontal axis interval of the second feature classification, j is the counting index, z is the number of the second target classification, is the length of the horizontal axis interval of the jth second target classification, The second coefficient for the jth second target classification.

[0048] Preferably, the step of translating the non-deformed area comprises the following steps:

[0049] The non-deformed area is filled into the scaled deformed area to obtain a corrected image, wherein the filling satisfies: the edge of the non-deformed area coincides with the edge of the deformed area, and the relative position of the grid in the non-deformed area in the first corrected image is consistent with the relative position in the corrected image.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] By classifying and identifying the actual imaging pictures, and performing linearization and scaling processing based on the identification results, when the demonstration system composed of the vehicle-mounted projection equipment encounters geometric distortion, it can autonomously identify and make targeted adjustments to the distorted parts of the actual imaging pictures. During the adjustment, it performs adaptive stretching according to the image ratio to ensure that the stretched part does not appear unbalanced, and performs translation adjustment on the non-deformed part, thereby ensuring the integrity of the overall image display. The entire process is implemented through algorithms, which can save time and improve adjustment speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of a method for controlling the image correction of the projection quality of a vehicle-mounted projection device according to the present invention;

[0053] Figure 2 A schematic diagram of a process of establishing a corresponding relationship between grids in an actual imaging picture and a normal imaging picture according to the present invention;

[0054] Figure 3 A schematic diagram of a process of comparing an actual imaging picture with a normal imaging picture to obtain at least one curved edge grid according to the present invention;

[0055] Figure 4 A schematic diagram of a process of performing coordinate linearization processing on a curved edge grid according to the present invention, wherein the actual imaging picture is corrected to a first corrected image;

[0056] Figure 5 A schematic diagram of a process of comparing a first corrected image with a normal imaging picture to obtain at least one deformed mesh according to the present invention;

[0057] Figure 6 It is a schematic diagram of the process of scaling the deformation area according to the present invention. DETAILED DESCRIPTION

[0058] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0059] Reference Figure 1 As shown, a method for controlling the image correction of the projection quality of a vehicle-mounted projection device includes:

[0060] Forming a sample grid image for correction testing, and obtaining a normal imaging picture of the sample grid image on the projection screen during normal projection;

[0061] During actual projection, the vehicle-mounted projection device uses the sample grid image for projection, captures the projection image, and obtains the actual imaging image;

[0062] The projection screen is modeled with coordinates, with the lower left corner of the projection screen as the origin, the horizontal direction as the horizontal axis, and the vertical direction as the vertical axis, to obtain the coordinates of the pixel points in the actual imaging picture and the normal imaging picture;

[0063] Recording a grid in the actual imaging picture as a first grid, recording a grid in the normal imaging picture as a second grid, establishing a corresponding relationship between the grids in the actual imaging picture and the grids in the normal imaging picture, comparing the actual imaging picture with the normal imaging picture, and obtaining at least one curved edge grid, where the curved edge grid is the grid in the actual imaging picture;

[0064] The curved edge mesh is subjected to coordinate linearization processing, the actual imaging picture is corrected into a first corrected image, and the first corrected image is compared with the normal imaging picture to obtain at least one deformed mesh;

[0065] Summarize the deformed meshes to obtain the deformed regions, and classify the deformed meshes in two ways: the deformed meshes with the same abscissa at the center of the deformed regions are classified into the first deformed mesh classification, and the deformed meshes with the same ordinate at the center of the deformed regions are classified into the second deformed mesh classification;

[0066] The area other than the deformed area in the first corrected image is regarded as the non-deformed area;

[0067] The deformed area is scaled and the non-deformed area is translated to complete the image correction.

[0068] When the projection screen is used for a long time, various bends, protrusions or depressions are prone to appear on its surface, which will cause local distortion of the picture. Therefore, the image is processed at this location to reduce the degree of distortion. In this solution, the actual imaging picture of the sample grid image is compared with the normal imaging picture to obtain a corresponding adjustment solution. During actual projection, this solution is used to adjust the picture. Since the problems of the projection screen are almost the same in a short period of time, this solution can be used for projection within this period of time. When this period of time is exceeded, the actual imaging picture of the sample grid image is compared with the normal imaging picture again to obtain an updated adjustment solution. During actual projection, the updated adjustment solution is used to adjust the picture.

[0069] Reference Figure 2 As shown, establishing a correspondence between the grids in the actual imaging picture and the normal imaging picture includes the following steps:

[0070] In the actual imaging picture, the number of first grids on the left side of the first grid is counted as the first horizontal number, and the number of first grids on the lower side of the first grid is counted as the first vertical number;

[0071] In a normal imaging picture, the number of second grids on the left side of the second grid is counted as the second horizontal number, and the number of second grids on the lower side of the second grid is counted as the second vertical number;

[0072] A corresponding relationship is established between a first grid and a second grid in which the first horizontal number is equal to the second horizontal number and the first vertical number is equal to the second vertical number.

[0073] Although the grid distribution in the actual imaging screen and the normal imaging screen will be deformed, their relative positions will not change. Therefore, the position of the grid can be described by the number of grids on the left and the number of grids on the bottom of each grid. When the relative positions of the grids in the actual imaging screen and the normal imaging screen are consistent, the number of grids on the left and the number of grids on the bottom are consistent.

[0074] Reference Figure 3 As shown, comparing the actual imaging picture with the normal imaging picture to obtain at least one curved edge grid includes the following steps:

[0075] Take at least one first sampling point at the edge of the first grid, perform fitting on the first sampling point, and obtain a first fitting function, where the first fitting function is a function expression of the edge of the first grid;

[0076] Take at least one second sampling point at the edge of the second grid, perform fitting on the second sampling point, and obtain a second fitting function, where the second fitting function is a function expression of the edge of the second grid;

[0077] Using the first curvature integral formula, calculate the first integrated curvature of the first grid;

[0078] Using the second curvature integral formula, calculate the second integrated curvature of the second grid;

[0079] When the difference between the first comprehensive curvature and the second comprehensive curvature of the corresponding first grid and second grid is greater than a preset value, the first grid is used as a curved edge grid;

[0080] The first curvature integral formula is as follows: ,

[0081] Among them, the integral path is the path corresponding to the edge of the first grid, A is the first comprehensive curvature, is the second-order derivative of the first fitting function, is the first derivative of the first fitting function, and x is the abscissa of a point on the edge of the first grid;

[0082] The second curvature integral formula is as follows: ,

[0083] Among them, the integral path is the path corresponding to the edge of the second grid, B is the second comprehensive curvature, is the second derivative of the second fitting function, is the first derivative of the second fitting function, and t is the abscissa of the point on the edge of the second grid.

[0084] Taking the edge of the first grid as an example, It is the curvature of the point on the edge of the first grid. The curvature describes the bending degree of the edge of the first grid. The path integral is performed to obtain the first comprehensive curvature. Similarly, the second comprehensive curvature is obtained. When the edge of the first grid is not distorted, the difference between the first comprehensive curvature and the second comprehensive curvature should be within the allowable range. Otherwise, it means that the edge of the first grid is distorted.

[0085] Reference Figure 4 As shown, the coordinate linearization process is performed on the curved edge grid, and the actual imaging picture is corrected to the first corrected image, including the following steps:

[0086] Obtain the first grid whose center coordinates have the smallest difference with the center of the curved edge grid as the first characteristic grid;

[0087] Obtaining the difference between the ordinate of a point on the upper edge of the first feature grid and the ordinate of the center of the first feature grid as the first distance;

[0088] Obtain the difference between the abscissa of a point on the left edge of the first feature grid and the abscissa of the center of the first feature grid as the second distance;

[0089] The coordinates of the points in the curved edge grid are linearized using a linearization formula, and the actual imaging image is corrected to a first corrected image;

[0090] The linearization formula is as follows: ,

[0091] Among them, m and n are the horizontal and vertical coordinates of the center of the curved edge grid, k and l are the horizontal and vertical coordinates of the points in the curved edge grid, respectively. is the first fitting function, for is the inverse function of the linearized curved edge mesh, where D is the first distance, d is the second distance, and K and L are the horizontal and vertical coordinates of the points in the linearized curved edge mesh, respectively.

[0092] During the linearization process, the coordinates of the points in the curved edge mesh are transformed. The transformation is adjusted according to the proportion of the curved edge mesh edge being stretched into a straight line, rather than just linearizing its edge. All points in the curved edge mesh are adjusted accordingly, thereby ensuring the coordination of the image in the curved edge mesh.

[0093] Reference Figure 5 As shown, comparing the first corrected image with the normal imaging picture to obtain at least one deformed mesh includes the following steps:

[0094] The grid in the first corrected image is used as the third grid, and the grid correspondence relationship of the first corrected image before correction is used to correspond the third grid to the second grid;

[0095] The area of ​​the third grid is compared with the area of ​​the corresponding second grid. When the difference between the area of ​​the third grid and the area of ​​the corresponding second grid is greater than a preset value, the third grid is used as a deformable grid.

[0096] After linearization, some grids are inconsistent in size with the grids in normal projection, so these grids need to be identified and stretched, but the stretching of grids in different positions is different. In this scheme, two stretching adjustments are made, namely horizontal and vertical adjustments. During the adjustment, since the grids are very small, the grids in the same row horizontally can use the same stretching ratio, and the grids in the same column vertically can also use the same stretching ratio. After stretching, the error generated is within the allowable range, but when stretching, there is a sequence of grid stretching. Therefore, it is necessary to consider the coordinate changes of the grid that is stretched first and incorporate them into the coordinates of the grid that is transformed later.

[0097] Reference Figure 6 As shown, scaling the deformation area includes the following steps:

[0098] Forming a ordinate interval for the first deformed mesh classification, wherein the ordinate interval is a value range of the ordinates of the deformed meshes in the first deformed mesh classification;

[0099] The ratio of the side length of the upper side of the deformed mesh in the first deformed mesh classification to the side length of the upper side of the corresponding second mesh is used as the first coefficient;

[0100] forming a horizontal coordinate interval for the second deformed mesh classification, wherein the horizontal coordinate interval is a value range of the horizontal coordinates of the deformed meshes in the second deformed mesh classification;

[0101] The ratio of the side length of the left side of the deformed mesh in the second deformed mesh classification to the side length of the left side of the corresponding second mesh is used as the second coefficient;

[0102] Randomly select a point in the deformation area as the deformation point, obtain the first deformed mesh classification located on the left side of the deformation point as the first target classification, and obtain the first deformed mesh classification containing the deformation point as the first feature classification;

[0103] Acquire a second deformed mesh classification located below the deformation point as a second target classification, and acquire a second deformed mesh classification including the deformation point as a second feature classification;

[0104] Use the scaling formula to calculate the scaled coordinates of the deformation point. When scaling, scale the deformation point to the scaled coordinates of the deformation point.

[0105] The scaling formula is as follows: ,

[0106] Among them, O is the ordinate of the scaled coordinate of the deformation point, o is the ordinate of the deformation point, is the first coefficient of the first feature classification, s is the left endpoint of the ordinate interval of the first feature classification, i is the counting index, r is the number of the first target classification, is the length of the ordinate interval of the i-th first target classification, is the first coefficient of the i-th first target classification, P is the abscissa of the scaled coordinates of the deformation point, and p is the abscissa of the deformation point. is the second coefficient of the second feature classification, u is the left endpoint of the horizontal axis interval of the second feature classification, j is the counting index, z is the number of the second target classification, is the length of the horizontal axis interval of the jth second target classification, The second coefficient for the jth second target classification.

[0107] The translation of the non-deformed area includes the following steps:

[0108] The non-deformed area is filled into the scaled deformed area to obtain a corrected image, wherein the filling satisfies: the edge of the non-deformed area coincides with the edge of the deformed area, and the relative position of the grid in the non-deformed area in the first corrected image is consistent with the relative position in the corrected image.

[0109] Here, the relative position consistency means that the number of grids located on the left and below the same grid in the first modified image is consistent with the number of grids located on the left and below the same grid in the corrected image.

[0110] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, the above-mentioned image correction control method applied to the projection quality of the vehicle-mounted projection device is executed.

[0111] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0112] To sum up, the advantages of the present invention are: by classifying and identifying the actual imaging picture, and performing linearization and scaling processing according to the identification results, when the demonstration system composed of the vehicle-mounted projection equipment encounters geometric distortion, it can perform autonomous identification and make targeted adjustments to the distorted parts of the actual imaging picture. During the adjustment, adaptive stretching is performed according to the image ratio to ensure that the stretched part will not appear unbalanced. The part that has not been deformed is translated and adjusted, thereby ensuring the integrity of the overall image display. The whole process is implemented through an algorithm, which can save time and improve the adjustment speed.

[0113] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for controlling the image correction of the projection quality of a vehicle-mounted projection device, characterized in that: include: Forming a sample grid image for correction testing, and obtaining a normal imaging picture of the sample grid image on the projection screen during normal projection; During actual projection, the vehicle-mounted projection device uses the sample grid image for projection, captures the projection image, and obtains the actual imaging image; The projection screen is modeled with coordinates, with the lower left corner of the projection screen as the origin, the horizontal direction as the horizontal axis, and the vertical direction as the vertical axis, to obtain the coordinates of the pixel points in the actual imaging picture and the normal imaging picture; Recording a grid in the actual imaging picture as a first grid, recording a grid in the normal imaging picture as a second grid, establishing a corresponding relationship between the grids in the actual imaging picture and the grids in the normal imaging picture, comparing the actual imaging picture with the normal imaging picture, and obtaining at least one curved edge grid, where the curved edge grid is the grid in the actual imaging picture; The curved edge mesh is subjected to coordinate linearization processing, the actual imaging picture is corrected into a first corrected image, and the first corrected image is compared with the normal imaging picture to obtain at least one deformed mesh; Summarize the deformed meshes to obtain the deformed regions, and classify the deformed meshes in two ways: the deformed meshes with the same abscissa at the center of the deformed regions are classified into the first deformed mesh classification, and the deformed meshes with the same ordinate at the center of the deformed regions are classified into the second deformed mesh classification; The area other than the deformed area in the first corrected image is regarded as the non-deformed area; The deformed area is scaled and the non-deformed area is translated to complete the image correction; The coordinate linearization process of the curved edge grid is performed, and the actual imaging picture is corrected into a first corrected image, which comprises the following steps: Obtain the first grid whose center coordinates have the smallest difference with the center of the curved edge grid as the first characteristic grid; Obtaining the difference between the ordinate of a point on the upper edge of the first feature grid and the ordinate of the center of the first feature grid as the first distance; Obtain the difference between the abscissa of a point on the left edge of the first feature grid and the abscissa of the center of the first feature grid as the second distance; The coordinates of the points in the curved edge grid are linearized using a linearization formula, and the actual imaging image is corrected to a first corrected image; The linearization formula is as follows: , Among them, m and n are the horizontal and vertical coordinates of the center of the curved edge grid, k and l are the horizontal and vertical coordinates of the points in the curved edge grid, respectively. is the first fitting function, for The inverse function of , D is the first distance, d is the second distance, K and L are the horizontal and vertical coordinates of the point in the curved edge mesh after linearization respectively; The step of comparing the first corrected image with the normal imaging picture to obtain at least one deformed mesh comprises the following steps: The grid in the first corrected image is used as the third grid, and the grid correspondence relationship of the first corrected image before correction is used to correspond the third grid to the second grid; Comparing the area of ​​the third grid with the area of ​​the corresponding second grid, when the difference between the area of ​​the third grid and the area of ​​the corresponding second grid is greater than a preset value, taking the third grid as a deformable grid; The scaling of the deformation area comprises the following steps: Forming a ordinate interval for the first deformed mesh classification, wherein the ordinate interval is a value range of the ordinates of the deformed meshes in the first deformed mesh classification; The ratio of the side length of the upper side of the deformed mesh in the first deformed mesh classification to the side length of the upper side of the corresponding second mesh is used as the first coefficient; forming a horizontal coordinate interval for the second deformed mesh classification, wherein the horizontal coordinate interval is a value range of the horizontal coordinates of the deformed meshes in the second deformed mesh classification; The ratio of the side length of the left side of the deformed mesh in the second deformed mesh classification to the side length of the left side of the corresponding second mesh is used as the second coefficient; Randomly select a point in the deformation area as the deformation point, obtain the first deformed mesh classification located on the left side of the deformation point as the first target classification, and obtain the first deformed mesh classification containing the deformation point as the first feature classification; Acquire a second deformed mesh classification located below the deformation point as a second target classification, and acquire a second deformed mesh classification including the deformation point as a second feature classification; Use the scaling formula to calculate the scaled coordinates of the deformation point. When scaling, scale the deformation point to the scaled coordinates of the deformation point. The scaling formula is as follows: , Among them, O is the ordinate of the scaled coordinate of the deformation point, o is the ordinate of the deformation point, is the first coefficient of the first feature classification, s is the left endpoint of the ordinate interval of the first feature classification, i is the counting index, r is the number of the first target classification, is the length of the ordinate interval of the i-th first target classification, is the first coefficient of the i-th first target classification, P is the abscissa of the scaled coordinates of the deformation point, and p is the abscissa of the deformation point. is the second coefficient of the second feature classification, u is the left endpoint of the horizontal axis interval of the second feature classification, j is the counting index, z is the number of the second target classification, is the length of the horizontal axis interval of the jth second target classification, The second coefficient for the j-th second target classification.

2. The image correction control method for the projection quality of a vehicle-mounted projection device according to claim 1 is characterized in that: The establishing of a correspondence between the grids in the actual imaging picture and the normal imaging picture comprises the following steps: In the actual imaging picture, the number of first grids on the left side of the first grid is counted as the first horizontal number, and the number of first grids on the lower side of the first grid is counted as the first vertical number; In a normal imaging picture, the number of second grids on the left side of the second grid is counted as the second horizontal number, and the number of second grids on the lower side of the second grid is counted as the second vertical number; A corresponding relationship is established between a first grid and a second grid in which the first horizontal number is equal to the second horizontal number and the first vertical number is equal to the second vertical number.

3. The method for controlling the image correction of the projection quality of a vehicle-mounted projection device according to claim 2, characterized in that: The comparing the actual imaging picture with the normal imaging picture to obtain at least one curved edge mesh comprises the following steps: Take at least one first sampling point at the edge of the first grid, perform fitting on the first sampling point, and obtain a first fitting function, where the first fitting function is a function expression of the edge of the first grid; Take at least one second sampling point at the edge of the second grid, perform fitting on the second sampling point, and obtain a second fitting function, where the second fitting function is a function expression of the edge of the second grid; Using the first curvature integral formula, calculate the first integrated curvature of the first grid; Using the second curvature integral formula, calculate the second integrated curvature of the second grid; When the difference between the first comprehensive curvature and the second comprehensive curvature of the corresponding first grid and second grid is greater than a preset value, the first grid is used as a curved edge grid; The first curvature integral formula is as follows: , Among them, the integral path is the path corresponding to the edge of the first grid, A is the first comprehensive curvature, is the second-order derivative of the first fitting function, is the first derivative of the first fitting function, and x is the abscissa of a point on the edge of the first grid; The second curvature integral formula is as follows: , Among them, the integral path is the path corresponding to the edge of the second grid, B is the second comprehensive curvature, is the second derivative of the second fitting function, is the first derivative of the second fitting function, and t is the abscissa of the point on the edge of the second grid.

4. The method for controlling the image correction of the projection quality of a vehicle-mounted projection device according to claim 3, characterized in that: The translation change of the non-deformed area comprises the following steps: The non-deformed area is filled into the scaled deformed area to obtain a corrected image, wherein the filling satisfies: the edge of the non-deformed area coincides with the edge of the deformed area, and the relative position of the grid in the non-deformed area in the first corrected image is consistent with the relative position in the corrected image.

Images