Infrared irradiation point position identification method, system and computer program

Abstract

The invention discloses an infrared irradiation point position identification method and system, and a computer program. The infrared irradiation point position identification method comprises the following steps: obtaining distortion parameters of an infrared camera; obtaining a homography matrix of the infrared camera for plane calibration; irradiating the infrared light spot on a screen to obtain a screen picture shot by the infrared camera; preprocessing the screen picture to obtain a black-and-white picture; taking a white area in the black-and-white image as an infrared irradiation point, and obtaining coordinates of the infrared irradiation point; and according to the distortion parameter and the homography matrix, converting the coordinate of the infrared irradiation point into the actual coordinate position of the infrared irradiation point on the screen plane. According to the technical scheme provided by the invention, the picture is subjected to black and white processing, interference and noise are removed, and coordinate transformation is carried out by using the camera distortion parameter and the homography matrix, so that accurate actual coordinates can be obtained; and an automatic identification and calibration means is adopted, so that the operation is convenient, the calibration precision is higher, and the ambient light interference resistance is high.

Description

Technical Field

[0001] The present invention relates to the field of image recognition, and in particular to a method, system and computer program for recognizing the position of an infrared irradiation point. Background Art

[0002] At present, the technical solution for obtaining the coordinates of the irradiation point of infrared rays on the TV screen has the disadvantages of low recognition accuracy, poor resistance to ambient light interference, and difficulty in manual calibration. It usually requires manual calibration by clicking on the screen, which is cumbersome to operate. In addition, the accuracy will be reduced due to errors in manual clicking. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the current technology, the present invention provides a method for identifying the position of an infrared irradiation point. The coordinate position of the infrared irradiation point is obtained by performing black and white processing on the photos taken by the camera, and then the coordinate transformation is performed using the camera distortion parameters and the homography matrix of the plane calibration to obtain the accurate actual coordinate position.

[0004] To achieve the above object, the embodiments of the present invention adopt the following technical solutions: A method for identifying the position of an infrared irradiation point comprises the following steps: Get the distortion parameters of the infrared camera; Get the homography matrix of the infrared camera shooting the screen for plane calibration; Irradiate the infrared light spot on the screen to obtain the screen picture taken by the infrared camera; Preprocess the screen image to obtain a black and white image; The white area in the black-and-white image is used as the infrared irradiation point, and the coordinates of the infrared irradiation point are obtained; According to the distortion parameters and the homography matrix, the coordinates of the infrared irradiation point are transformed into the actual coordinate position of the infrared irradiation point on the screen plane.

[0005] According to one aspect of the present invention, obtaining the distortion parameters of the infrared camera includes: Use an infrared camera to take multiple chessboard pictures; Get the coordinate position of each chessboard intersection in the chessboard image; According to the coordinate positions of the chessboard intersections in the chessboard image and the original coordinate positions, the convex optimization algorithm is used to minimize the likelihood value to obtain the distortion parameters of the infrared camera.

[0006] According to one aspect of the present invention, obtaining a homography matrix for plane calibration by photographing a screen with an infrared camera includes: Display a checkerboard on the screen; Use an infrared camera to take a picture of the chessboard on the screen; Get the coordinates of the intersection points of the chessboard in the screen chessboard image; According to the distortion parameters, the coordinates of the chessboard intersections are transformed into the original coordinates; According to the chessboard, the actual coordinates of the intersection points of the chessboard in the screen plane are reconstructed; According to the original coordinates and the actual coordinates, the homography matrix transformed from the original coordinates to the actual coordinates is obtained.

[0007] According to one aspect of the present invention, obtaining a homography matrix transforming the original coordinates to the actual coordinates according to the original coordinates and the actual coordinates includes: Using the four sets of original coordinates and actual coordinates, the transformation equations of the homography matrix are constructed; Solve the transformation equations and obtain all parameters of the homography matrix.

[0008] According to one aspect of the present invention, obtaining a homography matrix transforming the original coordinates to the actual coordinates according to the original coordinates and the actual coordinates includes: By using more than 4 sets of original coordinates and actual coordinates, the error value is minimized based on the convex optimization method to obtain all parameters of the homography matrix.

[0009] According to one aspect of the present invention, the preprocessing of the screen image to obtain the black and white image includes: Convert the screen image to black and white at the preset brightness dividing line; Blacken the area outside the screen in a black and white image; Based on the fuzzy algorithm, the black and white image is denoised so that large white areas in the black and white image can be connected together.

[0010] According to one aspect of the present invention, taking the white area in the black-and-white image as the infrared irradiation point and obtaining the coordinates of the infrared irradiation point comprises: Get the area of ​​each white area in the black and white image; Compare the size of each white area, and take the largest white area as the area where the infrared irradiation point is located; Get the smallest circle that can just fit the largest white area, and use the coordinates of the center of the circle as the coordinates of the infrared irradiation point.

[0011] According to one aspect of the present invention, the step of transforming the coordinates of the infrared irradiation point into the actual coordinate position of the infrared irradiation point on the screen plane according to the distortion parameter and the homography matrix comprises: According to the distortion parameter and the coordinates of the infrared irradiation point, the original coordinates of the infrared irradiation point before distortion are obtained; According to the homography matrix and the original coordinate position, the actual coordinate position of the infrared irradiation point in the screen plane is obtained.

[0012] An infrared irradiation point position recognition system, based on the infrared irradiation point position recognition method as described above, comprises: A distortion parameter acquisition module is used to obtain the distortion parameters of the infrared camera; A homography matrix acquisition module is used to obtain the homography matrix of the infrared camera shooting the screen for plane calibration; The shooting module is used to irradiate infrared light spots on the screen to obtain screen images taken by the infrared camera; A preprocessing module is used to preprocess the screen image to obtain a black and white image; A coordinate acquisition module, used to use the white area in the black-and-white image as the infrared irradiation point to obtain the coordinates of the infrared irradiation point; The coordinate transformation module is used to transform the coordinates of the infrared irradiation point into the actual coordinate position of the infrared irradiation point on the screen plane according to the distortion parameters and the homography matrix.

[0013] A computer program, which, when executed, implements the steps of the infrared irradiation point position recognition method as described above.

[0014] Advantages of the present invention: The present invention provides a method for identifying the position of an infrared irradiation point. The infrared irradiation point coordinate position is obtained by performing black-and-white processing on the photos taken by the camera, removing the interference area and the noise influence, and then using the camera distortion parameter and the homography matrix of the plane calibration to perform coordinate transformation, so as to obtain the accurate actual coordinate position. The method adopts the means of automatic identification and calibration, which is not only easy to operate, but also has higher calibration accuracy and strong resistance to ambient light interference.

[0015] The present invention also adopts a convex optimization algorithm to minimize the error value and obtain a more accurate homography matrix, so that the coordinate position of the infrared irradiation point finally obtained is more accurate; a mask is made to shield the part outside the screen in the camera shooting picture, which can reduce interference from the outside of the screen and improve the robustness of the recognition process. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 The present invention is a flowchart of a method for identifying the position of an infrared irradiation point. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Embodiment 1

[0020] like Figure 1 As shown, a method for identifying the position of an infrared irradiation point comprises the following steps: S1: Get the distortion parameters of the infrared camera.

[0021] Camera distortion is mainly caused by physical factors in the lens manufacturing process, which leads to image distortion. Distortion mainly includes radial distortion and tangential distortion.

[0022] Radial distortion is mainly caused by the irregular change of the radial curvature of the lens. This distortion will cause the straight lines in the image to become curved away from the center of the image. distorted , y distorted ) and the undistorted normalized image coordinates (x, y) can be expressed as:

[0023] Among them, r represents the distance from the image pixel to the center point of the image, and the calculation formula is r 2 = x 2 + y 2 ; k1, k2, and k3 are radial distortion coefficients, which determine the degree and shape of distortion. For lenses with less distortion, usually only k1 and k2 are needed to describe the distortion well; while for lenses with greater distortion, such as fisheye lenses, k3 coefficient is needed to describe the distortion more accurately.

[0024] Tangential distortion is mainly caused by the error in the assembly between the lens and the photosensitive element. This distortion will cause the straight lines in the image to tilt near the edge of the image. Tangential distortion can be described by the following formula:

[0025] Among them, p1 and p2 are the tangential distortion correction coefficients, which determine the degree and direction of tangential distortion.

[0026] Combining radial distortion and tangential distortion, we can get a comprehensive model of camera distortion:

[0027] Step S1 includes: S11: Use an infrared camera to take multiple chessboard pictures.

[0028] In practical applications, a calibration plate with known geometric shape and size is generally used to obtain the distortion parameters of the camera, usually a chessboard or dot array. The chessboard is generally a crisscross pattern of black and white blocks, and obvious intersections can be obtained as calibration feature points for positioning, etc.

[0029] S12: Obtain the coordinate position of each chessboard intersection in the chessboard image.

[0030] In practical applications, image processing software (such as OpenCV) can be used to extract calibration feature points, i.e., chessboard intersection points, from each chessboard image.

[0031] S13: According to the coordinate positions of the chessboard intersections in the chessboard image and the original coordinate positions, a convex optimization algorithm is used to minimize the likelihood value to obtain the distortion parameters of the infrared camera.

[0032] Using the comprehensive model of camera distortion, the distortion parameters of the infrared camera can be obtained based on the coordinates in the captured image (i.e., the distorted coordinates) and the original coordinates (the coordinates without distortion). The convex optimization algorithm is used to minimize the likelihood value, and the distortion parameters with the smallest error are obtained as much as possible to improve the coordinate accuracy.

[0033] S2: Obtain the homography matrix of the infrared camera shooting the screen for plane calibration.

[0034] Since the image captured by the camera is a projection of the three-dimensional space onto a two-dimensional plane, some spatial information will be lost, so it is necessary to calibrate the plane where the screen is located.

[0035] After removing the camera distortion, the camera image is approximated as a small hole model, and the points in space are projected into the image through the projection matrix. Since the projected points and the original points are on the same plane, a homography relationship is satisfied:

[0036] Among them, (x i , y i ) is the coordinate of the original point; ( , ) is the projection point coordinate; H is the homography matrix; s i is the scaling parameter, s i It can be calculated by the homography matrix. Therefore, the coordinate change relationship between the projection point and the original point should be theoretically: S i = h 31 xi + h 32 y i + h 33

[0037] Step S2 includes: S21: Display a checkerboard on the screen.

[0038] S22: Use an infrared camera to take a picture of the chessboard on the screen.

[0039] S23: Obtain the coordinates of the intersection points of the chessboard in the screen chessboard image.

[0040] S24: transforming the coordinates of the chessboard intersections into original coordinates according to the distortion parameters.

[0041] S25: Reconstruct the actual coordinates of the intersection points of the chessboard in the screen plane according to the chessboard.

[0042] S26: According to the original coordinates and the actual coordinates, a homography matrix transforming the original coordinates to the actual coordinates is obtained.

[0043] Step S26 includes: Using the four sets of original coordinates and actual coordinates, the transformation equations of the homography matrix are constructed; Solve the transformation equations and obtain all parameters of the homography matrix.

[0044] Using the above-mentioned coordinate change relationship between the projection point and the original point, we can construct a system of equations, solve the unknowns in the system of equations, and obtain all the parameters of the homography matrix.

[0045] S3: Irradiate the infrared light spot on the screen to obtain the screen image taken by the infrared camera.

[0046] S4: Preprocess the screen image to obtain a black and white image.

[0047] Step S4 includes: S41: Convert the screen image to black and white at the preset brightness dividing line.

[0048] In actual applications, the preset brightness dividing line may be 240, thereby converting the screen image into a black and white image.

[0049] S42: Blacken the area outside the screen in the black and white image.

[0050] Using the distortion parameters and homography matrix obtained in the previous steps, we only need to apply these two matrices to all pixels captured by the camera to determine whether the pixels are within the screen range. The pixels that are not within the screen range will be turned black to prevent the area outside the screen from interfering with the subsequent infrared irradiation points.

[0051] S43: Based on the fuzzy algorithm, the black and white image is denoised so that large white areas in the black and white image can be connected together.

[0052] S5: Taking the white area in the black-and-white image as the infrared irradiation point, and obtaining the coordinates of the infrared irradiation point.

[0053] Step S5 includes: S51: Obtain the area of ​​each white region in the black and white image.

[0054] S52: Compare the sizes of the white areas, and use the white area with the largest area as the area where the infrared irradiation point is located.

[0055] S53: Obtain the smallest circle that can just fit the largest white area, and use the coordinates of the center of the circle as the coordinates of the infrared irradiation point.

[0056] S6: According to the distortion parameter and the homography matrix, the coordinates of the infrared irradiation point are transformed into the actual coordinate position of the infrared irradiation point on the screen plane.

[0057] Step S6 includes: S61: Acquire the original coordinates of the infrared irradiation point before distortion according to the distortion parameters and the coordinates of the infrared irradiation point.

[0058] S62: According to the homography matrix and the original coordinate position, the actual coordinate position of the infrared irradiation point in the screen plane is obtained.

[0059] By using the distortion parameters and homography matrix obtained in the previous step, the coordinate change relationship between the distortion comprehensive model and the homography matrix is ​​inversely calculated, and the precise actual coordinate position of the infrared irradiation point in the screen plane can be obtained.

[0060] In practical applications, the present method can reduce the error of dozens of pixels in existing solutions to an average of 5 pixels.

[0061] The beneficial effect of this embodiment is that the method obtains the coordinate position of the infrared irradiation point by performing black-and-white processing on the photos taken by the camera, removing the interference area and the noise influence, and then uses the camera distortion parameters and the homography matrix of the plane calibration to perform coordinate transformation, so as to obtain the accurate actual coordinate position. The method adopts the means of automatic recognition and calibration, which is not only easy to operate, but also has higher calibration accuracy and strong resistance to ambient light interference.

[0062] Embodiment 2 like Figure 1 As shown, a method for identifying the position of an infrared irradiation point comprises the following steps: S1: Get the distortion parameters of the infrared camera.

[0063] Camera distortion is mainly caused by physical factors in the lens manufacturing process, which leads to image distortion. Distortion mainly includes radial distortion and tangential distortion.

[0064] Radial distortion is mainly caused by the irregular change of the radial curvature of the lens. This distortion will cause the straight lines in the image to become curved away from the center of the image. distorted , y distorted ) and the undistorted normalized image coordinates (x, y) can be expressed as:

[0065] Among them, r represents the distance from the image pixel to the center point of the image, and the calculation formula is r 2 = x 2 + y 2 ; k1, k2, and k3 are radial distortion coefficients, which determine the degree and shape of distortion. For lenses with less distortion, usually only k1 and k2 are needed to describe the distortion well; while for lenses with greater distortion, such as fisheye lenses, k3 coefficient is needed to describe the distortion more accurately.

[0066] Tangential distortion is mainly caused by the error in the assembly between the lens and the photosensitive element. This distortion will cause the straight lines in the image to tilt near the edge of the image. Tangential distortion can be described by the following formula:

[0067] Among them, p1 and p2 are the tangential distortion correction coefficients, which determine the degree and direction of tangential distortion.

[0068] Combining radial distortion and tangential distortion, we can get a comprehensive model of camera distortion:

[0069] Step S1 includes: S11: Use an infrared camera to take multiple chessboard pictures.

[0070] In practical applications, a calibration plate with known geometric shape and size is generally used to obtain the distortion parameters of the camera, usually a chessboard or dot array. The chessboard is generally a crisscross pattern of black and white blocks, and obvious intersections can be obtained as calibration feature points for positioning, etc.

[0071] S12: Obtain the coordinate position of each chessboard intersection in the chessboard image.

[0072] In practical applications, image processing software (such as OpenCV) can be used to extract calibration feature points, i.e., chessboard intersection points, from each chessboard image.

[0073] S13: According to the coordinate positions of the chessboard intersections in the chessboard image and the original coordinate positions, a convex optimization algorithm is used to minimize the likelihood value to obtain the distortion parameters of the infrared camera.

[0074] Using the comprehensive model of camera distortion, the distortion parameters of the infrared camera can be obtained based on the coordinates in the captured image (i.e., the distorted coordinates) and the original coordinates (the coordinates without distortion). The convex optimization algorithm is used to minimize the likelihood value, and the distortion parameters with the smallest error are obtained as much as possible to improve the coordinate accuracy.

[0075] S2: Obtain the homography matrix of the infrared camera shooting the screen for plane calibration.

[0076] Since the image captured by the camera is a projection of the three-dimensional space onto a two-dimensional plane, some spatial information will be lost, so it is necessary to calibrate the plane where the screen is located.

[0077] After removing the camera distortion, the camera image is approximated as a small hole model, and the points in space are projected into the image through the projection matrix. Since the projected points and the original points are on the same plane, a homography relationship is satisfied:

[0078] Among them, r represents the distance from the image pixel to the center point of the image, and the calculation formula is r 2 = x 2 + y 2 ; k1, k2, and k3 are radial distortion coefficients, which determine the degree and shape of distortion. For lenses with less distortion, usually only k1 and k2 are needed to describe the distortion well; while for lenses with greater distortion, such as fisheye lenses, k3 coefficient is needed to describe the distortion more accurately.

[0079] Tangential distortion is mainly caused by the error in the assembly between the lens and the photosensitive element. This distortion will cause the straight lines in the image to tilt near the edge of the image. Tangential distortion can be described by the following formula: Among them, (x i , y i ) is the coordinate of the original point; ( , ) is the projection point coordinate; H is the homography matrix; si is the scaling parameter, s i It can be calculated by the homography matrix. Therefore, the coordinate change relationship between the projection point and the original point should be theoretically: S i = h 31 x i + h 32 y i + h 33

[0080]

[0081] Step S2 includes: S21: Display a checkerboard on the screen.

[0082] S22: Use an infrared camera to take a picture of the chessboard on the screen.

[0083] S23: Obtain the coordinates of the intersection points of the chessboard in the screen chessboard image.

[0084] S24: transforming the coordinates of the chessboard intersections into original coordinates according to the distortion parameters.

[0085] S25: Reconstruct the actual coordinates of the intersection points of the chessboard in the screen plane according to the chessboard.

[0086] S26: According to the original coordinates and the actual coordinates, a homography matrix transforming the original coordinates to the actual coordinates is obtained.

[0087] Step S26 includes: By using more than 4 sets of original coordinates and actual coordinates, the error value is minimized based on the convex optimization method to obtain all parameters of the homography matrix.

[0088] The total error ε is:

[0089] By using the convex optimization method and minimizing the total error ε, more accurate homography matrix parameters can be obtained. The convex optimization algorithm can avoid the situation where the error of a single point causes the entire screen plane to deviate when the number of points is too small, making the final result inaccurate. The adaptive convex optimization algorithm can automatically exclude points with too large errors from participating in the calculation, and finally obtain a plane that is closest to the actual result.

[0090] S3: Irradiate the infrared light spot on the screen to obtain the screen image taken by the infrared camera.

[0091] S4: Preprocess the screen image to obtain a black and white image.

[0092] Step S4 includes: S41: Convert the screen image to black and white at the preset brightness dividing line.

[0093] In actual applications, the preset brightness dividing line may be 240, thereby converting the screen image into a black and white image.

[0094] S42: Blacken the area outside the screen in the black and white image.

[0095] In practical applications, since there may be interference from outside the screen in the camera image, in order to improve the robustness of the recognition process, this method creates a mask to shield the part outside the screen in the camera image.

[0096] Using the distortion parameters and homography matrix obtained in the previous step, these two matrices are applied to all pixels captured by the camera to determine whether the pixel is within the screen range, and the mask can be calculated. Pixels that are not within the screen range are turned black to prevent the area outside the screen from interfering with subsequent infrared irradiation points.

[0097] Traverse all the pixels of the black and white image, calculate the points projected onto the screen plane, and use the mask to blacken the points whose coordinates are beyond the screen range.

[0098] S43: Based on the fuzzy algorithm, the black and white image is denoised so that large white areas in the black and white image can be connected together.

[0099] S5: Taking the white area in the black-and-white image as the infrared irradiation point, and obtaining the coordinates of the infrared irradiation point.

[0100] Step S5 includes: S51: Obtain the area of ​​each white region in the black and white image.

[0101] S52: Compare the sizes of the white areas, and use the white area with the largest area as the area where the infrared irradiation point is located.

[0102] S53: Obtain the smallest circle that can just fit the largest white area, and use the coordinates of the center of the circle as the coordinates of the infrared irradiation point.

[0103] S6: According to the distortion parameter and the homography matrix, the coordinates of the infrared irradiation point are transformed into the actual coordinate position of the infrared irradiation point on the screen plane.

[0104] Step S6 includes: S61: Acquire the original coordinates of the infrared irradiation point before distortion according to the distortion parameters and the coordinates of the infrared irradiation point.

[0105] S62: According to the homography matrix and the original coordinate position, the actual coordinate position of the infrared irradiation point in the screen plane is obtained.

[0106] By using the distortion parameters and homography matrix obtained in the previous step, the coordinate change relationship between the distortion comprehensive model and the homography matrix is ​​inversely calculated, and the precise actual coordinate position of the infrared irradiation point in the screen plane can be obtained.

[0107] In practical applications, the present method can reduce the error of dozens of pixels in existing solutions to an average of 5 pixels.

[0108] The beneficial effects of this embodiment are: this method also uses a convex optimization algorithm to minimize the error value and obtain a more accurate homography matrix, so that the coordinate position of the infrared irradiation point finally obtained is more accurate; making a mask to shield the part outside the screen in the camera shooting picture can reduce interference from the outside of the screen and improve the robustness of the recognition process.

[0109] Embodiment 3 An infrared irradiation point position recognition system, based on the infrared irradiation point position recognition method as described in Embodiment 1 or 2, comprises: A distortion parameter acquisition module is used to obtain the distortion parameters of the infrared camera; A homography matrix acquisition module is used to obtain the homography matrix of the infrared camera shooting the screen for plane calibration; The shooting module is used to irradiate infrared light spots on the screen to obtain screen images taken by the infrared camera; A preprocessing module is used to preprocess the screen image to obtain a black and white image; A coordinate acquisition module, used to use the white area in the black-and-white image as the infrared irradiation point to obtain the coordinates of the infrared irradiation point; The coordinate transformation module is used to transform the coordinates of the infrared irradiation point into the actual coordinate position of the infrared irradiation point on the screen plane according to the distortion parameters and the homography matrix.

[0110] Embodiment 4 A computer program, when executed, implements the steps of the infrared irradiation point position recognition method as described in embodiment one or two.

[0111] Embodiment 5 A readable storage medium stores a computer program as described in Example 4, and when the computer program is executed, the steps of the infrared irradiation point position recognition method as described in Example 1 or 2 are implemented.

[0112] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the art within the technical scope disclosed in the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for identifying the position of an infrared irradiation point, characterized in that: The following steps are involved: Get the distortion parameters of the infrared camera; Get the homography matrix of the infrared camera shooting the screen for plane calibration; Irradiate the infrared light spot on the screen to obtain the screen picture taken by the infrared camera; Preprocess the screen image to obtain a black and white image; The white area in the black-and-white image is used as the infrared irradiation point, and the coordinates of the infrared irradiation point are obtained; According to the distortion parameters and the homography matrix, the coordinates of the infrared irradiation point are transformed into the actual coordinate position of the infrared irradiation point on the screen plane.

2. The infrared irradiation point position recognition method according to claim 1, characterized in that: The step of obtaining the distortion parameters of the infrared camera includes: Use an infrared camera to take multiple chessboard pictures; Get the coordinate position of each chessboard intersection in the chessboard image; According to the coordinate positions of the chessboard intersections in the chessboard image and the original coordinate positions, the convex optimization algorithm is used to minimize the likelihood value to obtain the distortion parameters of the infrared camera.

3. The infrared irradiation point position recognition method according to claim 1, characterized in that: The method of obtaining the homography matrix of the infrared camera shooting the screen for plane calibration includes: Display a checkerboard on the screen; Use an infrared camera to take a picture of the chessboard on the screen; Get the coordinates of the intersection points of the chessboard in the screen chessboard image; According to the distortion parameters, the coordinates of the chessboard intersections are transformed into the original coordinates; According to the chessboard, the actual coordinates of the intersection points of the chessboard in the screen plane are reconstructed; According to the original coordinates and the actual coordinates, the homography matrix transformed from the original coordinates to the actual coordinates is obtained.

4. The infrared irradiation point position recognition method according to claim 3, characterized in that: The step of obtaining a homography matrix for transforming the original coordinates to the actual coordinates according to the original coordinates and the actual coordinates includes: Using the four sets of original coordinates and actual coordinates, the transformation equations of the homography matrix are constructed; Solve the transformation equations and obtain all parameters of the homography matrix.

5. The infrared irradiation point position recognition method according to claim 3, characterized in that: The step of obtaining a homography matrix for transforming the original coordinates to the actual coordinates according to the original coordinates and the actual coordinates includes: By using more than 4 sets of original coordinates and actual coordinates, the error value is minimized based on the convex optimization method to obtain all parameters of the homography matrix.

6. The infrared irradiation point position recognition method according to claim 1, characterized in that: The preprocessing of the screen image to obtain the black and white image includes: Convert the screen image to black and white at the preset brightness dividing line; Blacken the area outside the screen in a black and white image; Based on the fuzzy algorithm, the black and white image is denoised so that large white areas in the black and white image can be connected together.

7. The infrared irradiation point position recognition method according to claim 1, characterized in that: The method of using the white area in the black-and-white image as the infrared irradiation point and obtaining the coordinates of the infrared irradiation point includes: Get the area of ​​each white area in the black and white image; Compare the size of each white area, and take the largest white area as the area where the infrared irradiation point is located; Get the smallest circle that can just fit the largest white area, and use the coordinates of the center of the circle as the coordinates of the infrared irradiation point.

8. The infrared irradiation point position recognition method according to claim 1, characterized in that: The step of transforming the coordinates of the infrared irradiation point into the actual coordinate position of the infrared irradiation point on the screen plane according to the distortion parameter and the homography matrix comprises: According to the distortion parameter and the coordinates of the infrared irradiation point, the original coordinates of the infrared irradiation point before distortion are obtained; According to the homography matrix and the original coordinate position, the actual coordinate position of the infrared irradiation point in the screen plane is obtained.

9. An infrared irradiation point position recognition system, characterized in that: The infrared irradiation point position recognition method according to any one of claims 1 to 8 comprises: A distortion parameter acquisition module is used to obtain the distortion parameters of the infrared camera; A homography matrix acquisition module is used to obtain the homography matrix of the infrared camera shooting the screen for plane calibration; The shooting module is used to irradiate infrared light spots on the screen to obtain screen images taken by the infrared camera; A preprocessing module is used to preprocess the screen image to obtain a black and white image; A coordinate acquisition module, used to use the white area in the black-and-white image as the infrared irradiation point to obtain the coordinates of the infrared irradiation point; The coordinate transformation module is used to transform the coordinates of the infrared irradiation point into the actual coordinate position of the infrared irradiation point on the screen plane according to the distortion parameters and the homography matrix.

10. A computer program, characterized in that When the computer program is executed, the steps of the infrared irradiation point position recognition method according to any one of claims 1 to 8 are implemented.

Images