A method for calibrating geometric parameters of a focused light field camera

Abstract

The present application belongs to the technical field of camera calibration, and discloses a focusing light field camera geometric parameter calibration method, which comprises the following steps: using a focusing light field camera to collect and acquire a white image and a corresponding checkerboard original light field image under different object distances; determining the focal plane of the camera based on the original light field image; using an edge diffusion function information fitting-based corner point detection algorithm to calculate the checkerboard corner point coordinate information in the original light field image on the focal plane; calculating the center coordinate information of each microlens in the microlens array of the camera and the size information of a single microlens in the microlens array based on the white image; calculating the pixel block size in the microlens sub-image taken by the center sub-aperture image according to the above information; and calculating the distance from the virtual imaging surface to the microlens array and the distance from the microlens array to the plane where the photosensitive element is located based on the pixel block size and in combination with a Gaussian imaging formula. The present application can improve the calibration accuracy of the geometric parameters of the focusing light field camera.

Description

Technical Field

[0001] This invention belongs to the field of camera calibration technology, and more specifically, relates to a method for calibrating the geometric parameters of a focusing light field camera. Background Technology

[0002] Traditional camera imaging directly integrates the light rays in a scene, and its depth of field is limited by the camera aperture. To obtain an image with a large depth of field, the camera aperture must be reduced, which leads to a decrease in camera resolution and a loss of image signal-to-noise ratio. Light field cameras can increase the depth of field while eliminating the mechanical focusing process, saving imaging time. However, in the application of light field cameras, because the distance between some optical components is very short, the accuracy of the results is greatly affected by manufacturing and assembly errors. Therefore, the light field camera needs to be calibrated before use.

[0003] For the calibration of focusing light field cameras, the traditional method is to extract the center sub-aperture image of the original light field image and then calibrate the light field camera according to Zhang Zhengyou's calibration method. However, the center sub-aperture image of the original light field image needs to be calibrated before it can be more accurate, so the existing methods have low calibration accuracy. Summary of the Invention

[0004] This invention provides a method for calibrating the geometric parameters of a focusing light field camera, thereby solving the problem of low calibration accuracy of focusing light field cameras in the prior art.

[0005] This invention provides a method for calibrating the geometric parameters of a focusing light field camera, comprising the following steps:

[0006] S1: Use a focusing light field camera to acquire the original light field images of the checkerboard pattern at different object distances; use the focusing light field camera to acquire a white image;

[0007] S2: Based on multiple original light field images obtained at different object distances, determine the focal plane of the focusing light field camera;

[0008] S3: Using a corner detection algorithm based on edge diffusion function information fitting, the coordinate information of the checkerboard corner points in the original light field image located on the focal plane is calculated; based on the white image, the center coordinate information of each microlens in the microlens array of the focusing light field camera and the size information of a single microlens in the microlens array are calculated.

[0009] S4: Based on the coordinate information of the corner points of the chessboard, the center coordinate information of each microlens, and the size information of a single microlens, calculate the pixel block size in the microlens sub-image of the central sub-aperture image;

[0010] S5: Based on the pixel block size and combined with the Gaussian imaging formula, calculate the distance from the virtual imaging surface to the microlens array and the distance from the microlens array to the plane where the photosensitive element is located.

[0011] Preferably, in step S1, the original light field image of the checkerboard includes: a first original light field image corresponding to the reference checkerboard calibration board and a second original light field image corresponding to the standard checkerboard calibration board.

[0012] Preferably, the reference checkerboard calibration plate and the standard checkerboard calibration plate are fixed on the same plane. By gradually moving the plane where the calibration plate is located, the distance between the plane where the calibration plate is located and the focusing light field camera is adjusted. After each movement of the same distance, a first light field original image and a second light field original image at that distance are acquired respectively. After multiple acquisitions at different object distances, the number of acquired first light field original images and second light field original images is the same.

[0013] Preferably, in step S1, the method of acquiring a white image using the focusing light field camera includes: fixing a white paper in front of the lens of the focusing light field camera, adjusting the aperture of the focusing light field camera so that the light spot appearing on the photosensitive element after the light passes through the microlens array of the focusing light field camera does not cause crosstalk, and taking a picture of the white paper to obtain the white image.

[0014] Preferably, in step S2, the method for determining the focal plane of the focusing light field camera includes: obtaining pixel information of the checkerboard edge after microlens imaging based on multiple original light field images obtained at different object distances; calculating the corner pixel size of the checkerboard at different object distances and the actual size represented by a single pixel based on the pixel information; and taking the object plane corresponding to the smallest corner size of the calculated standard checkerboard as the focal plane of the focusing light field camera.

[0015] The calculation method for the corner dimensions of the standard chessboard grid is as follows:

[0016]

[0017] In the formula, L a To calculate the corner dimensions of the standard chessboard grid, C r To reference the actual dimensions of the corner points of the chessboard, L r To reference the corner pixel dimensions in the original first light field image corresponding to the checkerboard calibration board, L s The pixel dimensions of the corner points in the original image of the second light field corresponding to the standard checkerboard calibration board.

[0018] Preferably, in step S3, the corner detection algorithm based on edge diffusion function information fitting is used to calculate the corner coordinate information of the chessboard grid, which includes the following sub-steps:

[0019] Obtain the Region of Interest (ROI) of the original light field image located at the focal plane;

[0020] In the ROI region, a mean filtering algorithm is applied to eliminate imaging noise, and a Hough line detection algorithm is applied to the image region corresponding to a single microlens to obtain corner edge lines;

[0021] Based on the center line of the corner edge line, multiple edge blocks located in different directions of the corner are obtained;

[0022] The point where the derivative is maximum is found by fitting the pixels in the edge block, and this point is taken as the edge point; the coordinates of the checkerboard corner points are obtained by calculating the average of the coordinates of the edge points in the horizontal and vertical directions.

[0023] Preferably, the coordinates of the corner points of the chessboard are calculated using the following formula:

[0024]

[0025] In the formula, x ui With y ui Let a and y represent the x and y coordinates of the checkerboard corner points in the image of the i-th microlens, respectively. ij b ij c ij With d ij x represents the edge points in the edge blocks of different directions in the i-th microlens sub-image, respectively. aij With x dij These are the x-coordinates and y-coordinates of the edge points in the upper and lower edge blocks, respectively. bij With y cij y = n1, n2, n3, and n4 are the ordinates of the edge points in the left and right edge blocks, respectively, and n1, n2, n3, and n4 are the number of edge points in the four edge blocks surrounding the corner point of the chessboard grid.

[0026] Preferably, in step S3, the center coordinate information of each microlens in the microlens array and the size information of a single microlens in the microlens array are obtained in the following way: the white image is binarized after being low-pass filtered, and the centroid of the connected region is calculated to obtain the center coordinates of each microlens in the microlens array and the diameter of a single microlens in the microlens array.

[0027] Preferably, in step S4, the pixel block size is calculated as follows:

[0028]

[0029] In the formula, p is the pixel block size, ui u represents the coordinates of the image formed by the virtual imaging point after passing through the i-th microlens in the microlens array. i+1 M represents the coordinates of the image formed by the virtual imaging point after passing through the (i+1)th microlens in the microlens array. i M represents the center coordinates of the i-th microlens in the microlens array. i+1 δ represents the center coordinates of the (i+1)th microlens in the microlens array, D represents the diameter of a single microlens in the microlens array, δ represents the size of a single pixel, and n is the number of checkerboard corner points after imaging through the microlens array.

[0030] Preferably, in step S5, the distance from the virtual imaging surface to the microlens array and the distance from the microlens array to the plane where the photosensitive element is located are obtained by simultaneously solving the following formulas:

[0031]

[0032]

[0033] In the formula, f represents the focal length of the microlens array, p is the pixel block size, D represents the diameter of a single microlens in the microlens array, and δ represents the size of a single pixel.

[0034] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0035] This invention proposes a geometric parameter calibration method for a focusing light field camera that differs from existing technologies. First, the focusing light field camera acquires original light field images of a checkerboard pattern at different object distances, and then acquires a white image. Next, based on multiple original light field images obtained at different object distances, the focal plane of the focusing light field camera is determined. Then, a corner detection algorithm based on edge spread function information fitting is used to calculate the coordinates of the checkerboard corner points in the original light field image at the focal plane. Based on the white image, the center coordinates of each microlens in the microlens array of the focusing light field camera and the size of a single microlens in the microlens array are calculated. Then, based on the checkerboard corner coordinates, the center coordinates of each microlens, and the size of a single microlens, the pixel block size in the microlens sub-image taken from the central sub-aperture image is calculated. Finally, based on the pixel block size and combined with the Gaussian imaging formula, the distance from the virtual imaging surface to the microlens array and the distance from the microlens array to the plane where the photosensitive element is located are calculated. This invention utilizes the geometric relationships in the imaging process of a focusing light field camera and employs a precise corner detection algorithm, thereby improving the accuracy of geometric parameter calibration for focusing light field cameras. Attached Figure Description

[0036] Figure 1This is a schematic diagram illustrating the imaging principle of a focusing light field camera.

[0037] Figure 2 A schematic diagram of central sub-aperture image extraction for a focusing light field camera;

[0038] Figure 3 The experimental setup used for calibration is shown in the diagram; among which, Figure 3 (a) in the diagram is a physical image of the experimental setup. Figure 3 Image (b) is an image of the reference chessboard calibration board. Figure 3 (c) is an image of the standard chessboard calibration board. Detailed Implementation

[0039] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0040] This embodiment provides a method for calibrating the geometric parameters of a focusing light field camera. (See also...) Figures 1 to 3 This includes the following steps:

[0041] S1: Use a focusing light field camera 10 to acquire the original light field images of the checkerboard pattern corresponding to different object distances; use the focusing light field camera 10 to acquire white images.

[0042] Specifically, the reference checkerboard calibration plate 7 and the standard checkerboard calibration plate 8 are mounted on the precision electric stage 9. By moving the precision electric stage 9, the original light field images of the checkerboard are captured by the focusing light field camera 10 at different object distances.

[0043] The original light field image of the checkerboard includes: a first original light field image corresponding to the reference checkerboard calibration plate 7 and a second original light field image corresponding to the standard checkerboard calibration plate 8.

[0044] The reference checkerboard calibration plate 7 and the standard checkerboard calibration plate 8 are fixed on the same plane. By gradually moving the plane where the calibration plate is located, the distance between the plane where the calibration plate is located and the focusing light field camera 10 is adjusted. After each movement of the same distance, the first light field original image and the second light field original image at that distance are acquired respectively. After multiple acquisitions at different object distances, the number of acquired first light field original images and second light field original images is the same.

[0045] The method of acquiring a white image using the focusing light field camera 10 includes: fixing a white paper in front of the lens of the focusing light field camera 10, adjusting the aperture of the focusing light field camera 10 so that the light spot appearing on the photosensitive element 5 after the light passes through the microlens array 4 of the focusing light field camera 10 does not cause crosstalk, and taking a picture of the white paper to obtain the white image.

[0046] S2: Based on multiple original light field images obtained at different object distances, determine the focal plane of the focusing light field camera 10.

[0047] Specifically, the method for determining the focal plane of the focusing light field camera 10 includes: obtaining pixel information of the checkerboard edge after microlens imaging based on multiple original light field images obtained at different object distances; calculating the corner pixel size of the checkerboard at different object distances and the actual size represented by a single pixel based on the pixel information; and taking the object plane corresponding to the smallest corner size of the calculated standard checkerboard as the focal plane of the focusing light field camera 10.

[0048] The corner dimensions of the standard chessboard grid are calculated using the following formula:

[0049]

[0050] In the formula, L a To calculate the corner dimensions of the standard chessboard grid, C r To reference the actual dimensions of the corner points of the chessboard, L r To reference the corner pixel dimensions in the original first light field image corresponding to the checkerboard calibration board, L s The pixel dimensions of the corner points in the original image of the second light field corresponding to the standard checkerboard calibration board.

[0051] The difference between edge points is the corner point size, and the average value between edge points is the corner point coordinate. That is, the present invention uses the edge diffusion information of checkerboard corner points to detect and confirm the focal plane of the focusing light field camera 10.

[0052] S3: Using a corner detection algorithm based on edge diffusion function information fitting, the coordinate information of the checkerboard corner points in the original light field image located on the focal plane is calculated; based on the white image, the center coordinate information of each microlens in the microlens array 4 of the focusing light field camera 10 and the size information of a single microlens in the microlens array 4 are calculated.

[0053] Step S3, calculating the coordinates of the corner points of the chessboard grid, specifically includes the following sub-steps:

[0054] S31: Obtain the ROI region of the original light field image located at the focal plane.

[0055] S32: Apply a mean filtering algorithm to the ROI region to eliminate imaging noise, and apply the Hough line detection algorithm to the image region corresponding to a single microlens to obtain corner edge lines.

[0056] S33: Based on the center line of the corner edge line, obtain multiple edge blocks located in different directions of the corner.

[0057] S34: Find the point where the derivative is maximum by fitting the pixels in the edge block, and take this point as the edge point; obtain the coordinates of the checkerboard corner points by calculating the average of the coordinates of the edge points in the horizontal and vertical directions.

[0058] The coordinates of the chessboard corner points in S34 are calculated using the following formula:

[0059]

[0060] In the formula, x ui With y ui Let a and y represent the x and y coordinates of the checkerboard corner points in the image of the i-th microlens, respectively. ij b ij c ij With d ij x represents the edge points in the edge blocks of different directions in the i-th microlens sub-image, respectively. aij With x dij These are the x-coordinates and y-coordinates of the edge points in the upper and lower edge blocks, respectively. bij With y cij y = n1, n2, n3, and n4 are the ordinates of the edge points in the left and right edge blocks, respectively, and n1, n2, n3, and n4 are the number of edge points in the four edge blocks surrounding the corner point of the chessboard grid.

[0061] In step S3, the center coordinate information of each microlens in the microlens array 4 and the size information of a single microlens in the microlens array 4 are obtained in the following way: the white image is binarized after being low-pass filtered, and the centroid of the connected region is calculated to obtain the center coordinates of each microlens in the microlens array 4 and the diameter of a single microlens in the microlens array 4.

[0062] S4: Based on the coordinate information of the corner points of the chessboard, the center coordinate information of each microlens, and the size information of a single microlens, calculate the pixel block size in the microlens sub-image of the central sub-aperture image.

[0063] That is, the pixel block size is calculated based on the focusing imaging optical path of the focusing light field camera 10 and the coordinates of the checkerboard corner points obtained in S3.

[0064] The pixel block size is calculated as follows:

[0065]

[0066] In the formula, p is the pixel block size, u i u represents the coordinates of the image formed by the virtual imaging point after passing through the i-th microlens in the microlens array. i+1 M represents the coordinates of the image formed by the virtual imaging point after passing through the (i+1)th microlens in the microlens array. i M represents the center coordinates of the i-th microlens in the microlens array. i+1 δ represents the center coordinates of the (i+1)th microlens in the microlens array, D represents the diameter of a single microlens in the microlens array, δ represents the size of a single pixel, and n is the number of checkerboard corner points after imaging through the microlens array.

[0067] S5: Based on the pixel block size and combined with the Gaussian imaging formula, calculate the distance from the virtual imaging surface 6 to the microlens array 4 and the distance from the microlens array 4 to the plane where the photosensitive element 5 is located.

[0068] Specifically, the distance *a* from the virtual imaging surface 6 to the microlens array 4 and the distance *b* from the microlens array 4 to the plane where the photosensitive element 5 is located can be obtained by simultaneously solving the following formulas:

[0069]

[0070]

[0071] In the formula, f represents the focal length of the microlens array, which is a known parameter; p is the pixel block size; D represents the diameter of a single microlens in the microlens array; and δ represents the size of a single pixel.

[0072] The geometric parameters of the focusing light field camera 10 are calibrated by using the pixel block size p and the Gaussian imaging formula. The geometric parameters include the distance b from the microlens array 4 to the plane where the photosensitive element 5 is located (i.e., the detector plane) and the distance a from the virtual imaging surface 6 to the microlens array 4.

[0073] pass Figure 2 The geometric relationships in the extraction of the central sub-aperture image of the light field can be used to obtain the relationship between the pixel block size p and the parameters a and b. The pattern on the virtual imaging surface 6 is imaged on the photosensitive element 5. When the pixel block size p satisfies the equation... At this time, the rendered image achieves optimal matching. Since the distance b between the microlens array 4 and the plane where the photosensitive element 5 is located is fixed after packaging, the pixel block size p of the rendered image depends on the value of a.

[0074] After determining the focal plane of the focusing light field camera 10 and calculating the pixel size p in step S2, the imaging formula of the microlens is further combined: a and b can then be obtained.

[0075] The invention will now be further explained in conjunction with the imaging principle.

[0076] The imaging principle of the focusing light field camera 10 is as follows: Figure 1 As shown, the defocus plane 1 or the focusing plane 2 forms an image on the virtual imaging surface 6 through the main lens 3, namely points R and R1. The microlens array 4 performs secondary acquisition on the image formed by the virtual imaging surface 6, and finally forms sub-images corresponding to the distribution of microlenses on the photosensitive element (e.g., CCD) 5.

[0077] See Figure 1 , Figure 2 Taking the image formed on the virtual imaging surface 6 by the main lens 3 through the focusing plane 2 as an example, the main lens 3 can be regarded as a thin lens. When the object distance (i.e., the distance between the second object surface 2 and the main lens 3) is u, the object distance u and the image distance v (i.e., the distance between the main lens 3 and point R) satisfy the Gaussian imaging formula:

[0078]

[0079] Where F is the focal length of the main lens.

[0080] The image distance *v* changes with the object distance *u*, the distance *a* between the virtual imaging point *R* and the microlens array 4 changes with *u*, and the distance *b* between the microlens array 4 and the photosensitive element 5 changes with *a*. Treating the microlens array 4 as a thin lens for imaging, the virtual imaging point *R* is imaged on the photosensitive element 5 after passing through the microlens array 4, satisfying the Gaussian imaging formula:

[0081]

[0082] Where f is the focal length of the microlens array.

[0083] Based on the geometric relationship of the image formed after the same virtual imaging point passes through the microlens array 4, the relationship between a and b can be derived as follows:

[0084]

[0085] Among them, u i u represents the coordinates of the image formed by the virtual imaging point after passing through the i-th microlens in the microlens array. i+1 M represents the coordinates of the image formed by the virtual imaging point after passing through the (i+1)th microlens in the microlens array. iM represents the center coordinates of the i-th microlens in the microlens array. i+1 This represents the center coordinates of the (i+1)th microlens in the microlens array.

[0086] like Figure 2 The diagram shows the extraction of the central sub-aperture image from the focusing light field camera 10. The virtual imaging surface 6 is imaged on the photosensitive element 5 after passing through the microlens array 4. When the pixel block size p of the photosensitive element 5 satisfies equation (4), the central sub-aperture image reaches the optimal matching state. The following formula can be obtained from the geometric similarity relationship:

[0087]

[0088] Where D represents the diameter of a single microlens in the microlens array (i.e., the distance between two adjacent microlenses in the microlens array), δ represents the size of a single pixel, and p represents the size of a pixel block.

[0089] By combining equations (2), (3), and (4), the values ​​of the parameters a, b, and p to be calibrated can be obtained.

[0090] The invention will now be described in conjunction with the experimental setup.

[0091] See the diagram of the experimental setup used for calibration. Figure 3 In (a), the experimental setup includes a reference checkerboard calibration plate 7, a standard checkerboard calibration plate 8, a precision electric stage 9, a focusing light field camera 10, and a two-dimensional moving platform 11. Figure 3 Image (b) is an image of the reference chessboard calibration board 7. Figure 3 Image (c) in the figure represents the standard checkerboard calibration board 8. To avoid confusion between different corner points when capturing the checkerboard image, each of the two checkerboard calibration boards used in this invention has one corner point. The size C of the checkerboard corner point in the reference checkerboard calibration board 7 is... r =0.1mm.

[0092] The method for calibrating the geometric parameters of a focusing light field camera using the above experimental setup includes the following steps:

[0093] Step 1: Install and connect the experimental devices (including: install the reference checkerboard calibration plate 7 and the standard checkerboard calibration plate 8 on the precision electric stage 9, and install the focusing light field camera 10 on the two-dimensional moving platform 11), and adjust the two-dimensional moving platform 11 so that the checkerboard corner points of the reference checkerboard calibration plate 7 are aligned with the optical center of the main lens of the focusing light field camera 10;

[0094] Step 2: By adjusting the precision electric stage 9, the reference checkerboard calibration plate 7 is moved to gradually approach the focusing light field camera 10. The focusing light field camera 10 acquires the original light field image of the checkerboard after each adjustment and records the distance of each movement. More than 20 original light field images of the reference checkerboard calibration plate 7 are acquired at different distances. The more images acquired, the more accurate the calibration result.

[0095] Step 3: Move the two-dimensional calibration platform 11 so that the standard checkerboard calibration plate 8 is aligned with the optical center of the main lens of the focusing light field camera 10. Repeat the operation of step 2 and take multiple original light field images corresponding to the standard checkerboard calibration plate 8 at the same distance as in step 2.

[0096] Step 4: Replace the plane where the calibration plate is located with a piece of white paper. Adjust the aperture of the focusing light field camera 10 so that the light passing through the microlens array 4 appears as a small white dot on the photosensitive element 5 (i.e., the light spot does not cause crosstalk). Take a picture of the white paper to obtain a white image. After low-pass filtering, binarize the white image and calculate the centroid of the connected regions to obtain the center coordinates of each microlens in the microlens array 4 (the center coordinates of the i-th microlens are marked as M). i And the distance D between the centers of two adjacent microlenses in the microlens array 4;

[0097] Step 5: Calculate the size of the corner point at different object distances, and take the object plane corresponding to the smallest calculated corner point size of the standard checkerboard as the focal plane of the focusing light field camera 10;

[0098] Step 6: When at the focal plane position, accurately solve the coordinates u of the checkerboard corner points in different microlenses using the corner edge diffusion function information. i By combining equations (2), (3), and (4), the values ​​of the parameters a, b, and p to be calibrated can be obtained.

[0099] In summary, this invention, based on edge diffusion function information fitting, can accurately calculate the corner size and coordinates of a standard checkerboard pattern. The object plane corresponding to the minimum calculated corner size of the standard checkerboard pattern is taken as the focal plane of the focusing light field camera. Through the geometric relationships of optical imaging, the pixel size *p* in the microlens sub-image of the central sub-aperture image corresponding to the light field camera under the focal plane, and the relationship between the pixel size *p* and calibration parameters (including the distance *a* from the virtual imaging surface to the microlens array and the distance *b* from the microlens array to the plane where the photosensitive element is located) are derived. The geometric parameters of the focusing light field camera are then accurately calculated using the Gaussian imaging formula. In other words, this invention utilizes the edge diffusion information of the checkerboard corners to detect the focal plane and corner coordinates of the focusing light field camera, thus completing the calibration of the focusing light field camera. This method utilizes the geometric relationships of the imaging process of the focusing light field camera and employs a precise corner detection algorithm, improving the accuracy of the calibration results. It has practical significance and good application prospects.

[0100] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for calibrating geometry parameters of a focused light field camera, characterized in that, The method comprises the following steps: S1: acquiring original light field images of the chessboard under different object distances using a focusing light field camera; and acquiring a white image using the focusing light field camera; The original light field images of the chessboard comprise a first light field original image corresponding to a reference chessboard calibration board and a second light field original image corresponding to a standard chessboard calibration board; S2: determining a focal plane of the focusing light field camera based on the multiple original light field images obtained under different object distances; The implementation of determining the focal plane of the focusing light field camera comprises: obtaining pixel information of the edges of the chessboard after imaging by the microlenses based on the multiple original light field images obtained under different object distances; calculating the pixel size of the corner points of the chessboard corresponding to different object distances and the actual size represented by a single pixel according to the pixel information; and taking the object plane corresponding to the minimum size of the corner points of the standard chessboard as the focal plane of the focusing light field camera; The calculation of the size of the corner points of the standard chessboard is as follows: In the formula, is the calculated standard checkerboard corner point size, is the actual size of the corner point of the reference checkerboard, is the pixel size of the corner point in the first light field raw image corresponding to the reference checkerboard calibration board, is the pixel size of the corner point in the second light field raw image corresponding to the standard checkerboard calibration board; S3: calculating the coordinate information of the corner points of the chessboard in the original light field image at the focal plane by using an edge spread function information fitting-based corner point detection algorithm; and calculating the center coordinate information of each microlens in the microlens array of the focusing light field camera and the size information of a single microlens in the microlens array based on the white image; S4: calculating the size of the pixel block in the microlens sub-image taken by the center sub-aperture image according to the coordinate information of the corner points of the chessboard, the center coordinate information of each microlens, and the size information of a single microlens; S5: calculating the distance from the virtual imaging plane to the microlens array and the distance from the microlens array to the plane where the photosensitive element is located based on the size of the pixel block and in combination with a Gaussian imaging formula.

2. The focal-plane light-field camera geometry calibration method of claim 1, wherein, The reference chessboard calibration board and the standard chessboard calibration board are fixed on the same plane, and the distance between the plane where the calibration board is located and the focusing light field camera is adjusted by gradually moving the plane where the calibration board is located. After moving the same distance each time, the first light field original image and the second light field original image under the distance are respectively acquired; and the number of images of the first light field original image and the second light field original image obtained after multiple acquisitions under different object distances is the same.

3. The focal-plane light-field camera geometry calibration method of claim 1, wherein, In step S1, the implementation of acquiring the white image using the focusing light field camera comprises: fixing a white paper in front of the lens of the focusing light field camera, adjusting the aperture of the focusing light field camera so that the light spot presented on the photosensitive element after the light passes through the microlens array of the focusing light field camera does not cause crosstalk, and capturing the white paper to obtain the white image.

4. The focal-plane light-field camera geometry calibration method of claim 1, wherein, In step S3, the calculation of the coordinate information of the corner points of the chessboard by using the edge spread function information fitting-based corner point detection algorithm comprises the following sub-steps: Obtaining the ROI region of the original light field image at the focal plane; Applying a median filter algorithm to the ROI region to eliminate imaging noise, and applying a Hough line detection algorithm to the image region corresponding to a single microlens to obtain a corner edge line; Based on the center line of the corner edge line, a plurality of edge blocks in different directions of the corner point are obtained; A point with maximum derivative is found by fitting pixels in the edge block, and the point is taken as an edge point; Chessboard corner point coordinates are obtained by calculating average values of horizontal and vertical edge point coordinates.

5. The focal-plane light-field camera geometry calibration method of claim 4, wherein, The chessboard corner point coordinates are calculated by using the following formula: In the formula, x ui and y ui respectively represent the horizontal coordinate and the vertical coordinate of the chessboard corner point in the i-th microlens sub-image, a ij , b ij , c ij and d ij respectively represent the edge points in the edge blocks of different directions in the i-th microlens sub-image, x aij and x dij respectively represent the horizontal coordinates of the edge points in the upper and lower edge blocks, y bij and y cij respectively represent the vertical coordinates of the edge points in the left and right edge blocks, n 1, n 2, n 3, n 4 respectively represent the number of edge points in the four edge blocks around the chessboard corner point.

6. The focal-plane light-field camera geometry calibration method of claim 1, wherein, In step S3, the center coordinate information of each microlens in the microlens array and the size information of a single microlens in the microlens array are obtained by the following method: the white image is subjected to low-pass filtering and binarization processing, and the center of the connected region is calculated to obtain the center coordinate of each microlens in the microlens array and the diameter of a single microlens in the microlens array.

7. The focal-plane light-field camera geometry calibration method of claim 1, wherein, In step S4, the pixel block size is calculated as follows: where p is the size of the pixel block, u i represents the coordinate of the virtual imaging point after passing through the i-th micro-lens in the micro-lens array, u i+1 represents the coordinate of the virtual imaging point after passing through the i+1-th micro-lens in the micro-lens array, M i represents the coordinate of the center of the i-th micro-lens in the micro-lens array, M i+1 represents the coordinate of the center of the i+1-th micro-lens in the micro-lens array, D represents the diameter of a single micro-lens in the micro-lens array, δ represents the size of a single pixel, and n is the number of the checkerboard corner points after imaging through the micro-lens array.

8. The focal-plane light-field camera geometry calibration method of claim 1, wherein, In step S5, the distance from the virtual imaging plane to the microlens array and the distance from the microlens array to the plane where the photosensitive element is located are obtained by solving the following formula: In the formula, f represents the focal length of the microlens array, p is the pixel block size, D represents the diameter of a single microlens in the microlens array, and δ represents the size of a single pixel.

Images