Method for evaluating dynamic distortion based on view field angle

By fitting the field of view circle and calculating the field of view angle of pixel point positions, the problem of dynamic distortion in the VR module is solved, and the accuracy and efficiency of image processing are improved, and the shape of different field of view circles is adapted to.

CN118984376APending Publication Date: 2024-11-19ZHEJIANG SUNNY OPTICAL CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410997583.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The VR module has dynamic distortion during the projection of object space to the image plane, which is mainly caused by errors in optical lenses, display screen characteristics and eye tracking technology, resulting in inaccurate image distortion.

Method used

By fitting the field of view circle, calculate the field of view angle of the pixel point position, remove the central cross interference, automatically determine the origin, calculate the offset from the actual point to the ideal point, and improve accuracy and automation.

Benefits of technology

Accurate evaluation of dynamic distortion is achieved, the accuracy and efficiency of image processing is improved, and field circles of different sizes and shapes are adapted to field of view, simplifying the calculation process, and enhancing the reliability of subsequent processing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118984376A_ABST
    Figure CN118984376A_ABST
Patent Text Reader

Abstract

The invention provides a method for evaluating dynamic distortion based on a field angle, and the method comprises the following steps: extracting a central cross in an image, and calculating the deviation angle and distance between the center of the cross and the center of the image, and the projection in x and y directions; obtaining the center and radius of each field-of-view circle, and solving a field-of-view angle image of each pixel point in the target surface and field-of-view angle images projected in the x direction and the y direction; and calculating the offset of the actual point relative to the ideal point lattice diagram according to the lattice image.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method for evaluating dynamic distortion based on field of view angle. Background Art

[0002] A VR module is a virtual reality display device that can provide an immersive virtual environment, making users feel as if they are in the scene. In the process of projecting the object space to the image plane, the VR module often has dynamic distortion. The reasons for this include: the influence of optical lenses, such as the optical lenses in the VR module will cause light refraction, resulting in image distortion, especially when using lenses with a large field of view (FOV), the distortion is more obvious; the influence of the display screen, the characteristics of the display screen may also cause image distortion, for example, the resolution, refresh rate, pixel arrangement and other factors of the display screen may affect the image quality; errors in eye tracking technology: eye tracking technology is used to dynamically correct distortion, but existing eye tracking technology may have errors in accuracy and latency, resulting in the appearance of dynamic distortion; and head and eye movement: when the user moves his head or eyes, the degree of image distortion may change, which can also cause dynamic distortion. Summary of the invention

[0003] A major advantage of the present invention is that it provides a method for evaluating dynamic distortion based on field of view angle, wherein the method accurately calculates the field of view angle by fitting the field of view circle and considering the pixel position, thereby avoiding errors.

[0004] Another advantage of the present invention is that it provides a method for evaluating dynamic distortion based on field of view angle, wherein the method removes central cross interference, improves accuracy, and is adaptable to field of view circles of different sizes and shapes, and is universal.

[0005] Another advantage of the present invention is that it provides a method for evaluating dynamic distortion based on field of view angle, wherein the method uses the center of the cross as the origin, simplifies the calculation process, and improves efficiency.

[0006] Another advantage of the present invention is that it provides a method for evaluating dynamic distortion based on field of view angle, wherein the method calculates the offset from the actual point to the ideal point through precise position correspondence, thereby improving the accuracy and reliability of subsequent processing.

[0007] Another advantage of the present invention is that it provides a method for evaluating dynamic distortion based on field of view angle, wherein the method sorts the extracted actual coordinate points with the origin as the center, establishes an ideal point by calculating the average distance between the four upper, lower, left and right points closest to the origin, thereby ensuring the uniformity and rationality of the distribution of the ideal point.

[0008] Another advantage of the present invention is that it provides a method for evaluating dynamic distortion based on field of view angle, wherein the method can automatically determine the origin according to the position of the deviated pupil, thereby avoiding errors and uncertainties in manual selection and improving the degree of automation.

[0009] According to one aspect of the present invention, a method for evaluating dynamic distortion based on field of view angle of the present invention can achieve the above-mentioned objects and other objects and advantages, wherein the method for evaluating dynamic distortion comprises the following steps:

[0010] (1) Extract the central cross in the image and calculate the offset angle, distance, and projection in the x and y directions between the center of the cross and the center of the image;

[0011] (2) obtaining the center and radius of each field of view circle, and calculating the field of view angle image of each pixel point on the target surface and the field of view angle images projected in the x-direction and the y-direction; and

[0012] (3) Calculate the offset of the actual point relative to the ideal point dot matrix based on the dot matrix image.

[0013] According to one embodiment of the present invention, step (1) further comprises the following steps:

[0014] (1.1) Get the target surface area, invert the input image within the target surface area, and make the area of ​​interest the foreground. The specific formula is as follows: Value 目标 =255-Value 当前 ;

[0015] (1.2) Select an appropriate threshold to perform binary processing on the target area;

[0016] (1.3) Find the coordinates of the four endpoints of the cross at the center of the image, and find the straight lines where the horizontal and vertical line segments are located and the coordinates of the center of the cross (x cross ,y cross );as well as

[0017] (1.4) According to the distance formula, the distance L between the center of the cross and the center of the image, the offset angle θ, and the projections in the x and y directions are x' and y' respectively, where the calculation method is:

[0018] L=(x c -x cross ) 2 +(y c -y cross ) 2

[0019]

[0020] x′=sinθ

[0021] y′=cosθ

[0022] Among them, x c and c are the column and row indices of the center of the image.

[0023] According to an embodiment of the present invention, in the step (1.3), the rows of the thinned image are cut unchanged, and the columns are sub-images of 150 pixels on the left and right of the center point of the image. The sub-image pixels are traversed to find the coordinates of the top and bottom foreground pixels respectively, and converted into the coordinates of the image before cropping (x up ,y up ), (x bottom ,y bottom ); and the columns of the image after the refinement remain unchanged, and the columns are sub-images of 150 pixels above and below the center of the image. The sub-image pixels are traversed to find the coordinates of the leftmost and rightmost foreground pixels, and converted into the coordinates of the image before cropping (x left ,y left ), (x rigt ,y rigt ).

[0024] According to one embodiment of the present invention, the step (1.2) further comprises the following steps:

[0025] (1.2.1) Loop through all foreground pixels and mark as deleted the pixels that meet the following conditions:

[0026] 2<=N(P1)<=6, S(P1)=1, P2*P4*P6=0, P4*P6*P8=0; N(p1) represents the number of foreground pixels among the 8 pixels adjacent to P1, S(P1) represents the cumulative number of 0~1 appearing in P2~P9~P2 pixels, where 0 represents background and 1 represents foreground;

[0027] (1.2.2) Loop through all foreground pixels and mark pixels P1 that meet the following conditions as deleted:

[0028] x2<=N(P1)<=6, S(P1)=1, P2*P4*P8=0, P2*P6*P8=0; and

[0029] (1.2.3) Repeat steps (1.2.1) and (1.2.2) until no pixels are marked for deletion in the above two steps. The output result is the skeleton after the binary image is refined.

[0030] According to one embodiment of the present invention, the step (1.1) further comprises the following steps:

[0031] (1.1.3) Binarize the image and perform a closing operation on the image to filter out the influence of noise and preliminarily locate the target surface area; and

[0032] (1.1.4) Find all contours, take the area enclosed by the largest contour, select the appropriate corrosion factor to perform corrosion operation to obtain the target surface area.

[0033] According to one embodiment of the present invention, the step (2) further comprises the following steps:

[0034] (2.1) Using the least squares circle fitting method to fit different circles based on different point sets, the radius and center point of each circle are obtained;

[0035] (2.2) The average value of the fitted centers is the center of the concentric circles, that is, the centers and radii of multiple concentric circles are obtained; and

[0036] (2.3) The field of view angle of each pixel is calculated by interpolation operation according to the field of view angle of each field of view circle, and the field of view angle image radianImg is obtained. The field of view angles are projected in the X and Y directions as radianImgX and radianImgY respectively.

[0037] According to an embodiment of the present invention, in step (2.1), different circles are fitted using the least squares circle fitting method according to different point sets, thereby obtaining the radius and center point of each circle. The specific operation is as follows:

[0038] Let the center coordinates (h, k) and radius r. If there are n data points (x i ,y i );

[0039] The equation that describes the circle can be defined as: (xh) 2 +(yk) 2 =r 2 ;

[0040] Using the least squares method, it can be expressed as: minimize∑((x i -h) 2 +(y i -k) 2 -r 2 ) 2 ;

[0041] The optimization algorithm is used to finally find the optimal center coordinates (h, k) and radius r.

[0042] According to one embodiment of the present invention, the following steps are further included before step (2.1):

[0043] (2.0.1) Remove the center cross and calculate the distance from each foreground point to the two straight lines. If the distance from each foreground point to the vertical line and horizontal line is less than 50 and 25 pixels respectively, the foreground is set as the background.

[0044] (2.0.2) Calculate the distance from each foreground point in the image to the center of the cross; and

[0045] (2.0.3) Cluster different point sets according to the size of the distance.

[0046] According to one embodiment of the present invention, in the step (2.0.3), a point set is randomly initialized according to the distance from a certain point to the cross center, and each point is traversed; if the distance from a point to the cross center belongs to the existing set, it is added to the corresponding set; otherwise, a new set is created to store the point until all points are added to the corresponding set; if the point set is less than 100, the set point is deleted to avoid interference from stray points.

[0047] According to one embodiment of the present invention, step (3) further comprises the following steps:

[0048] (3.1) Extract the actual point coordinates of the input dot matrix image;

[0049] (3.2) extracting the values ​​of the actual point coordinates Pi (x, y) in the field of view angle image X and the field of view angle image Y, and calculating the point P (x, y) closest to the given pupil deviation point from all Pi, where the point P (x, y) is set as the origin O; and

[0050] (3.3) Calculate the ideal point spacing d, where the average distance from the origin O to the four points above, below, left and right is the ideal spacing; establish an ideal point vector map with the center point as the origin and d as the distance, and sort them according to the corresponding positions.

[0051] According to an embodiment of the present invention, the step (3) further comprises the following step: based on the correspondence between the ideal point and the actual point, drawing an arrow vector diagram pointing from the actual point to the ideal point.

[0052] According to one embodiment of the present invention, the step (3.1) further comprises the following steps:

[0053] (3.1.1) Binarize the input dot matrix image;

[0054] (3.1.2) Select a 5*5 structural element from the binarized image and perform a morphological closing operation on it; and

[0055] (3.1.3) The center point of each dot matrix is ​​extracted based on the minimum circumscribed rectangle of the contour, which is the proposed actual coordinate point position.

[0056] According to one embodiment of the present invention, the step (3.2) further comprises the following steps:

[0057] According to the actual point coordinates extracted through traversal, the x and y values ​​of the field of view angle projection in the image radianImg are taken out, and the distance between each projection point P(x, y) and the given pupil point is calculated, and the position of the projection point with the smallest distance is set as the origin O.

[0058] Further objects and advantages of the present invention will be fully apparent from an understanding of the following description and the accompanying drawings.

[0059] These and other objects, features and advantages of the present invention will be more fully understood from the following detailed description and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. In the accompanying drawings, unless otherwise specified, the same reference numerals are used to represent the same components. Among them:

[0061] Figure 1 is the original image before processing.

[0062] Figure 2 This is a thresholded image of a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention.

[0063] Figure 3 This is an image after the cross is removed according to a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention.

[0064] Figure 4 Detailed Description of the Invention The figure is an image of the field of view angle of a method for evaluating dynamic distortion based on the field of view angle according to a first preferred embodiment of the present invention.

[0065] Figure 5A and Figure 5B 1 is a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention, showing field of view angle images in the X and Y directions.

[0066] Figure 6 It is a schematic diagram of a local magnification of a dot matrix input into a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention.

[0067] Figure 7 It is a vector diagram from an ideal point to an actual point of a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention.

[0068] Figure 84 is a flow chart of a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention.

[0069] Fig. 9 4 is a schematic diagram of the steps of a method for evaluating dynamic distortion based on field of view angle according to a first preferred embodiment of the present invention. DETAILED DESCRIPTION

[0070] It should be pointed out that the embodiments shown in the drawings are only used as examples to specifically and vividly explain and illustrate the concept of the present invention. The size and structure thereof are not necessarily drawn to scale, nor do they constitute a limitation to the concept of the present invention.

[0071] The directional terms such as up, down, left, right, front, back, front, back, top, bottom, etc. mentioned or may be mentioned in this specification are defined relative to the structures shown in the various drawings. They are relative concepts and may change accordingly according to different positions and different usage conditions. Therefore, these or other directional terms should not be interpreted as restrictive terms.

[0072] It is to be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element may be one, while in another embodiment, the number of the element may be multiple, and the term "one" should not be understood as a limitation on the quantity.

[0073] Referring to the drawings of this application specification Figures 1 to 9 As shown, a method for evaluating dynamic distortion based on the field of view angle in accordance with the first preferred embodiment of the present application is explained in the following description. For the dynamic distortion of the VR module, the method for evaluating dynamic distortion is characterized by CRA (arc radius). Since the VR module needs to simulate the visual effect of the human eye, the method described in the present application is based on the image captured by the human eye lens. In order to evaluate the value of dynamic distortion, the method for evaluating dynamic distortion establishes a FOV (field of view angle) chart and a dot matrix chart, and converts the image height into the field of view angle of CRA through the mapping relationship between FOV and dot matrix. In this way, the dynamic distortion value of the VR module can be evaluated and corresponded to the angular resolution of the human eye to ensure that the displayed virtual environment is as close to the real world as possible. It is worth mentioning that the method for evaluating dynamic distortion can automatically and accurately calculate the actual deviation from the ideal based on the position of the deviated pupil.

[0074] Specifically, the method for evaluating dynamic distortion comprises the following steps:

[0075] (1) Extract the central cross in the image and calculate the offset angle, distance, and projection in the x and y directions between the center of the cross and the center of the image;

[0076] (2) obtaining the center and radius of each field of view circle, and calculating the field of view angle image of each pixel point on the target surface and the field of view angle images projected in the x-direction and the y-direction; and

[0077] (3) Calculate the offset of the actual point relative to the ideal point dot matrix based on the dot matrix image.

[0078] like Figure 1 As shown, the captured image, wherein the field of view angles of each field of view ring in the image from the inside to the outside are 5°, 10°, ..., 45°. In step (1) of the method for evaluating dynamic distortion described in the present application, the acquired image is subjected to average blur processing, and a 5x5 kernel is used to calculate the weighted average of each pixel neighborhood in the image, and the average value is replaced and output as the corresponding pixel of the image.

[0079] In the method for evaluating dynamic distortion, the step (1) further comprises the following steps:

[0080] (1.1) Get the target surface area, invert the input image within the target surface area, and make the area of ​​interest the foreground. The specific formula is as follows: Value 目标 =255-Value 当前 ;

[0081] (1.2) Select an appropriate threshold to perform binary processing on the target area;

[0082] (1.3) Find the coordinates of the four endpoints of the cross at the center of the image, and find the straight lines where the horizontal and vertical line segments are located and the coordinates of the center of the cross (x cross ,y cross );as well as

[0083] (1.4) According to the distance formula, the distance L between the center of the cross and the center of the image, the offset angle θ, and the projections in the x and y directions are x' and y' respectively, where the calculation method is:

[0084] L=(x c -x cross ) 2 +(y c -y cross ) 2

[0085]

[0086] x′=sinθ

[0087] y′=cosθ

[0088] Among them, x c and c are the column and row indices of the center of the image.

[0089] In the step (1.3) of the method for evaluating dynamic distortion, the rows of the refined image are cut unchanged, and the columns are sub-images of 150 pixels on the left and right of the center point of the image. The sub-image pixels are traversed to find the coordinates of the top and bottom foreground pixels respectively, and converted into the coordinates of the image before cropping (x up ,y up ), (x bottom ,y bottom ); and the columns of the image after the refinement remain unchanged, and the columns are sub-images of 150 pixels above and below the center of the image. The sub-image pixels are traversed to find the coordinates of the leftmost and rightmost foreground pixels, and converted into the coordinates of the image before cropping (x left ,y left ), (x rigt ,y rigt ).

[0090] It is worth mentioning that in the step (1.3) of the method for evaluating dynamic distortion, the straight line representation of the horizontal and vertical line segments and the cross center coordinates (x cross ,y cross ), the specific operations are as follows:

[0091] The slope of the line where the horizontal line segment lies is:

[0092] The straight line where the horizontal line segment lies is: y h =k h (x h -x left )+y left ;

[0093] The slope of the line where the vertical line segment lies is:

[0094] The line where the vertical line segment lies is: y v =k v (x v -x up )+y up ;

[0095] Calculate the intersection coordinates (x cross ,y cross ) is the cross center coordinate.

[0096] like Figure 2 As shown, a suitable threshold is selected to perform binarization processing on the target surface area. Specifically, the step (1.2) of the method for evaluating dynamic distortion further includes the following steps:

[0097] (1.2.1) Loop through all foreground pixels and mark as deleted the pixels that meet the following conditions:

[0098] 2<=N(P1)<=6, S(P1)=1, P2*P4*P6=0, P4*P6*P8=0; N(p1) represents the number of foreground pixels among the 8 pixels adjacent to P1, S(P1) represents the cumulative number of 0~1 appearing in P2~P9~P2 pixels, where 0 represents background and 1 represents foreground;

[0099] (1.2.2) Loop through all foreground pixels and mark pixels P1 that meet the following conditions as deleted:

[0100] x2<=N(P1)<=6, S(P1)=1, P2*P4*P8=0, P2*P6*P8=0; and

[0101] (1.2.3) Repeat steps (1.2.1) and (1.2.2) until no pixels are marked for deletion in the above two steps. The output result is the skeleton after the binary image is refined.

[0102] As an example, in a specific example of the present application, the complete pixel positions of P1 to P9 are as follows:

[0103] 9 2 3 8 1 4 7 6 5

[0104] The step (1.1) of the method for evaluating dynamic distortion further comprises the following steps:

[0105] (1.1.5) Binarize the image and perform a closing operation on the image to filter out the influence of noise and preliminarily locate the target surface area; and

[0106] (1.1.6) Find all contours, take the area enclosed by the largest contour, select the appropriate corrosion factor to perform corrosion operation to obtain the target surface area.

[0107] In a specific example of the present application, a triangular threshold is used to binarize an image. According to the image histogram, the maximum value of the longest distance from the maximum peak to the darkest end of the histogram is found, which is the threshold thresh. The binarization operation is performed using the following formula:

[0108]

[0109] To avoid interference, find all contours, take the area surrounded by the largest contour, select the appropriate 351*351 size corrosion factor for corrosion operation, and finally obtain the target surface area, which is the mask image.

[0110] In the preferred embodiment of the present application, the step (2) of the method for evaluating dynamic distortion further comprises the following steps:

[0111] (2.1) Using the least squares circle fitting method to fit different circles based on different point sets, the radius and center point of each circle are obtained;

[0112] (2.2) The average value of the fitted centers is the center of the concentric circles, that is, the centers and radii of multiple concentric circles are obtained; and

[0113] (2.3) The field of view angle of each pixel is calculated by interpolation operation according to the field of view angle of each field of view circle, and the field of view angle image radianImg is obtained. The field of view angles are projected in the X and Y directions as radianImgX and radianImgY respectively.

[0114] Specifically, in the step (2.1) of the method for evaluating dynamic distortion, different circles are fitted using the least squares circle fitting method according to different point sets, thereby obtaining the radius and center point of each circle. The specific operation is as follows:

[0115] Let the center coordinates (h, k) and radius r. If there are n data points (x i ,y i );

[0116] The equation that describes the circle can be defined as: (xh) 2 +(yk) 2 =r 2 ;

[0117] Using the least squares method, it can be expressed as: minimizeΣ((x i -h) 2 +(y i -k) 2 -r 2 ) 2 ;

[0118] The optimization algorithm is used to finally find the optimal center coordinates (h, k) and radius r.

[0119] As an example, in the step (2.2) of the method for evaluating dynamic distortion, the average value of the centers of the fitted circles is obtained as the center of the concentric circles, that is, the centers of 9 concentric circles and their radii are obtained, and the radii are sorted so that the radii from the inside to the outside are R[0], R[1], R[2], ..., R[8].

[0120] It is worth mentioning that the method for evaluating dynamic distortion in the preferred embodiment of the present application fits each field of view circle and takes into account the position of the pixel point, and then calculates the field of view angle according to the position of the pixel point on the field of view circle, and calculates the field of view angle of the projection of each pixel point to the x and y directions with the center of the cross as the origin, which is conducive to accurately calculating the field of view angle, avoiding errors, and improving the accuracy of the calculation.

[0121] like Figures 4 to 5B As shown, in the step (2.2) of the method for evaluating dynamic distortion, the field of view angle of each pixel is calculated by interpolation operation according to the field of view angle of each field of view circle, and the field of view angle image radianImg is obtained. The field of view angles projected in the X and Y directions are radianImgX and radianImgY respectively, as shown in FIG. Figure 5A and Figure 5B The specific operation (pseudo code) is shown in the following table:

[0122] Table 1 Algorithm for calculating the field of view angle of each pixel based on Visual Studio

[0123]

[0124]

[0125] The method for evaluating dynamic distortion further includes the following steps before step (2.1):

[0126] (2.0.1) Remove the center cross and calculate the distance from each foreground point to the two straight lines. If the distance from each foreground point to the vertical line and horizontal line is less than 50 and 25 pixels respectively, the foreground is set as the background.

[0127] (2.0.2) Calculate the distance from each foreground point in the image to the center of the cross; and

[0128] (2.0.3) Cluster different point sets according to the size of the distance.

[0129] like Figure 3 The figure shows a partial display of the image after removing the central cross. For the image obtained in step (1.2), the distance from each foreground point to the two straight lines is calculated respectively. If the distance from each foreground point to the vertical line and the horizontal line is less than 50 and 25 pixels respectively, the foreground is set as the background.

[0130] It is worth mentioning that since the center cross will have a certain impact on the extraction of the field of view circle, this method first removes the center cross, obtains the position of the center point of the cross, and then fits each field of view circle. Therefore, the method for evaluating dynamic distortion in this preferred embodiment of the present application improves accuracy by removing the interference of the center cross, and is adaptable to field of view circles of different sizes and shapes, and is universal.

[0131] In calculating the distance from each foreground point to the cross center in the image, assuming that the foreground point p p (x,y), then the distance from this point to the center of the cross is:

[0132]

[0133] In the step (2.0.3) of the method for evaluating dynamic distortion, a point set is randomly initialized according to the distance from a certain point to the center of the cross, and each point is traversed; if the distance from a point to the center of the cross belongs to the existing set, it is added to the corresponding set; otherwise, a new set is created to store the point until all points are added to the corresponding set; if the point set is less than 100, the set points are deleted to avoid interference from stray points.

[0134] It is worth mentioning that the method for evaluating dynamic distortion in the preferred embodiment of the present application first removes the central cross and obtains the position of the center point of the cross, then fits each field of view circle, and then calculates the field of view angle based on the position of the pixel point on the field of view circle, and calculates the field of view angle of the projection of each pixel point to the x and y directions with the center of the cross as the origin. That is, this method uses the center of the cross as the origin, simplifies the calculation process, and is conducive to improving efficiency.

[0135] The step (3) of the method for evaluating dynamic distortion in the preferred embodiment of the present application proposes a method for calculating the offset from the actual point to the ideal point based on the field of view angle, wherein the step (3) calculates the offset of the actual point relative to the ideal point dot matrix based on the dot matrix image. Due to the omission of the actual point extraction, the correspondence with the ideal point is inaccurate. Therefore, the method for evaluating dynamic distortion calculates the offset from the actual point to the ideal point through precise position correspondence, thereby improving the accuracy and reliability of subsequent processing.

[0136] The book search step (3) of the method for evaluating dynamic distortion in the preferred embodiment of the present application extracts the coordinates of each dot matrix by binarizing the input dot matrix image, and performs closing operations, deletes noise points and other operations, finds the projection value xy of the field of view angle of each point in the image in step 2 and calculates the point closest to the given eccentric pupil coordinate as the origin, and sorts the extracted actual coordinate points with the origin as the center; with the origin as the center, finds the average value of the distance between the four upper, lower, left and right points closest to the origin as the distance to establish an ideal point, and sorts and draws arrow vector diagrams from the ideal point to the actual point.

[0137] Specifically, the step (3) of the method for evaluating dynamic distortion in the preferred embodiment of the present application further includes the following steps:

[0138] (3.1) Extract the actual point coordinates of the input dot matrix image;

[0139] (3.2) extract the coordinates of the actual point in the field of view image X and the field of view image Y, i.e., Pi(x, y) (i=1, 2, ..., N), calculate the point P(x, y) that is closest to the given pupil deviation point from all Pi, and at this time, point P(x, y) is set as the origin O;

[0140] (3.3) Calculate the ideal point spacing d, where the average value of the distance from the origin O to the four points above, below, left and right is the ideal spacing, and establish an ideal point vector map with the center point as the origin and d as the distance, and sort them according to the corresponding positions.

[0141] It is worth mentioning that in this preferred embodiment of the present application, the method for evaluating dynamic distortion can automatically determine the origin according to the position of the deviated pupil, thereby avoiding the errors and uncertainties of manual selection and improving the degree of automation.

[0142] In this preferred embodiment of the present application, the method for evaluating dynamic distortion is to sort the extracted actual coordinate points with the origin as the center, and establish an ideal point by calculating the average distance between the four upper, lower, left and right points closest to the origin, thereby ensuring the uniformity and rationality of the distribution of the ideal point.

[0143] The method for evaluating dynamic distortion further includes the following steps between step (3.2) and step (3.3): starting from any point, find the four points closest to the point, and repeat this step until all points find the four points closest to the point; according to the relative position of the current point and its four nearest points, divide it into internal points, edge points and remaining coordinate points, and sort the internal points, edge points and remaining coordinate points.

[0144] Specifically, start from any point, calculate its distance to all other points, and find the 4 closest points. Repeat this step until all points have found the 4 closest points.

[0145] According to the relative position of the current point and its 4 nearest points, it is divided into internal points and edge points. The characteristic of internal points is that its 4 nearest points are roughly arranged in a diamond shape. Sort the internal points. First put all the internal points into a queue, then take one point from the queue as a reference point each time, and then find the other points closest to the point from the queue and sort them according to the relative position. Repeat this step until the queue is empty. Sort the edge points. For each edge point, find the closest internal point among the internal points and sort them according to the relative position. Sort the remaining coordinate points (i.e., unsorted coordinate points). First find the closest distance between these coordinate points and the edge points, and then sort them according to the distance and relative position. Sort according to the origin O. Start from the origin O and look for points in the same row / column in the direction of the edge, and sort these points (sort in ascending order from the center to the edge).

[0146] Furthermore, the step (3) of the method for evaluating dynamic distortion further comprises the following step: drawing an arrow vector diagram pointing from the actual point to the ideal point based on the correspondence between the ideal point and the actual point.

[0147] The step (3.1) of the method for evaluating dynamic distortion further comprises the following steps:

[0148] (3.1.1) Binarize the input dot matrix image;

[0149] (3.1.2) Select a 5*5 structural element from the binarized image and perform a morphological closing operation on it;

[0150] (3.1.3) The center point of each dot matrix is ​​extracted based on the minimum circumscribed rectangle of the contour, which is the proposed actual coordinate point position.

[0151] Furthermore, in the step (3.1) of the method for evaluating dynamic distortion, all contours appearing in the image are found, and contours with an area less than 4 are deleted to eliminate the influence of cluttered noise points.

[0152] The step (3.2) of the method for evaluating dynamic distortion further comprises the following steps:

[0153] According to the actual point coordinates extracted through traversal, the x and y values ​​of the field of view angle projection in the image radianImg are taken out, and the distance between each projection point P(x, y) and the given pupil point is calculated, and the position of the projection point with the smallest distance is set as the origin O.

[0154] It should be understood by those skilled in the art that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments, and the embodiments of the present invention may be deformed or modified in any way without departing from the principles.

[0155] The technical scope of the present invention is not limited to the contents in the above description. Those skilled in the art can make various deformations and modifications to the above embodiments without departing from the technical idea of ​​the present invention, and these deformations and modifications all belong to the protection scope of the present invention.

Claims

1. A method for evaluating dynamic distortion based on field of view angle, characterized in that: The method for evaluating dynamic distortion comprises the following steps: (1) Extract the central cross in the image and calculate the offset angle, distance, and projection in the x and y directions between the center of the cross and the center of the image; (2) obtaining the center and radius of each field of view circle, and calculating the field of view angle image of each pixel point on the target surface and the field of view angle images projected in the x-direction and the y-direction; as well as (3) Calculate the offset of the actual point relative to the ideal point dot matrix based on the dot matrix image.

2. The method for evaluating dynamic distortion according to claim 1, wherein the step (1) further comprises the following steps: (1.1) Get the target surface area, invert the input image within the target surface area, and make the area of ​​interest the foreground. The specific formula is as follows: Value 目标 =255-Value 当前 ; (1.2) Select an appropriate threshold to perform binary processing on the target area; (1.3) Find the coordinates of the four endpoints of the cross at the center of the image, and find the straight lines where the horizontal and vertical line segments are located and the coordinates of the center of the cross (x cross ,y cross );as well as (1.4) According to the distance formula, the distance L between the center of the cross and the center of the image, the offset angle θ, and the projections in the x and y directions are x' and y' respectively, where the calculation method is: L=(x c -x cross ) 2 +(and c -and cross ) 2 x′=sinθ y′=cosθ in, x c and c are the column and row indices of the center of the image.

3. The method for evaluating dynamic distortion according to claim 2, wherein in the step (1.3), the rows of the refined image are cut unchanged, and the columns are sub-images of 150 pixels on the left and right of the center point of the image, and the sub-image pixels are traversed to respectively find the coordinates of the top and bottom foreground pixels, and converted into the coordinates of the image before cropping (x up ,y up ), (x bottom ,y bottom ); and the columns of the image after the refinement remain unchanged, and the columns are sub-images of 150 pixels above and below the center of the image. The sub-image pixels are traversed to find the coordinates of the leftmost and rightmost foreground pixels, and converted into the coordinates of the image before cropping (x left ,y left ), (x rigt ,y rigt ).

4. The method for evaluating dynamic distortion according to claim 2, wherein the step (1.2) further comprises the following steps: (1.2.1) Loop through all foreground pixels and mark as deleted the pixels that meet the following conditions: 2<=N(P1)<=6, S(P1)=1, P2*P4*P6=0, P4*P6*P8=0; N(p1) represents the number of foreground pixels among the 8 pixels adjacent to P1, S(P1) represents the cumulative number of 0~1 appearing in P2~P9~P2 pixels, where 0 represents background and 1 represents foreground; (1.2.2) Loop through all foreground pixels and mark pixels P1 that meet the following conditions as deleted: x2<=N(P1)<=6, S(P1)=1, P2*P4*P8=0, P2*P6*P8=0; and (1.2.3) Repeat steps (1.2.1) and (1.2.2) until no pixels are marked for deletion in the above two steps. The output result is the skeleton after the binary image is refined.

5. The method for evaluating dynamic distortion according to claim 2, wherein the step (1.1) further comprises the following steps: (1.1.1) Binarize the image and perform a closing operation on the image to filter out the influence of noise and preliminarily locate the target area; and (1.1.2) Find all contours, take the area enclosed by the largest contour, select a suitable corrosion factor to perform corrosion operation to obtain the target surface area.

6. The method for evaluating dynamic distortion according to claim 4, wherein the step (2) further comprises the following steps: (2.1) Using the least squares circle fitting method to fit different circles based on different point sets, the radius and center point of each circle are obtained; (2.2) The average value of the fitted centers is the center of the concentric circles, that is, the centers and radii of multiple concentric circles are obtained; and (2.3) The field of view angle of each pixel is calculated by interpolation operation according to the field of view angle of each field of view circle, and the field of view angle image radianImg is obtained. The field of view angles are projected in the X and Y directions as radianImgX and radianImgY respectively.

7. The method for evaluating dynamic distortion according to claim 6, wherein in the step (2.1), different circles are fitted according to different point sets using the least squares circle fitting method, so as to obtain the radius and center point of each circle. The specific operation is as follows: Let the center coordinates (h, k) and radius r. If there are n data points (x i ,y i ); The equation that describes the circle can be defined as: (xh) 2 +(yk) 2 =r 2 ; Using the least squares method, it can be expressed as: minimize∑((x i -h) 2 +(y i -k) 2 -r 2 ) 2 ; The optimization algorithm is used to finally find the optimal center coordinates (h, k) and radius r.

8. The method for evaluating dynamic distortion according to claim 6, further comprising the following steps before step (2.1): (2.0.1) Remove the center cross and calculate the distance from each foreground point to the two straight lines. If the distance from each foreground point to the vertical line and horizontal line is less than 50 and 25 pixels respectively, the foreground is set as the background. (2.0.2) Calculate the distance from each foreground point in the image to the center of the cross; as well as (2.0.3) Cluster different point sets according to the size of the distance.

9. The method for evaluating dynamic distortion according to claim 8, wherein in the step (2.0.3), a point set is randomly initialized according to the distance from a certain point to the center of the cross, and each point is traversed; if the distance from a point to the center of the cross belongs to an existing set, it is added to the corresponding set; otherwise, a new set is created to store the point until all points are added to the corresponding set; if the point set is less than 100, the set point is deleted to avoid interference from stray points.

10. The method for evaluating dynamic distortion according to claim 6, wherein the step (3) further comprises the following steps: (3.1) Extract the actual point coordinates of the input dot matrix image; (3.2) extracting the values ​​of the actual point coordinates Pi (x, y) in the field of view angle image X and the field of view angle image Y, and calculating the point P (x, y) closest to the given pupil deviation point from all Pi, where the point P (x, y) is set as the origin O; and (3.3) Calculate the ideal point spacing d, where the average distance from the origin O to the four points above, below, left and right is the ideal spacing; establish an ideal point vector map with the center point as the origin and d as the distance, and sort them according to the corresponding positions.

11. The method for evaluating dynamic distortion according to claim 10, wherein the step (3) further comprises the following step: drawing an arrow vector diagram pointing from the actual point to the ideal point based on the correspondence between the ideal point and the actual point.

12. The method for evaluating dynamic distortion according to claim 11, wherein the step (3.1) further comprises the following steps: (3.1.1) Binarize the input dot matrix image; (3.1.2) Select a 5*5 structural element from the binarized image and perform a morphological closing operation on it; and (3.1.3) The center point of each dot matrix is ​​extracted based on the minimum circumscribed rectangle of the contour, which is the proposed actual coordinate point position.

13. The method for evaluating dynamic distortion according to claim 11, wherein the step (3.2) further comprises the following steps: According to the actual point coordinates extracted through traversal, the x and y values ​​of the field of view angle projection in the image radianImg are taken out, and the distance between each projection point P(x, y) and the given pupil point is calculated, and the position of the projection point with the smallest distance is set as the origin O.