A light field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration

Abstract

This application provides a light field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration, belonging to the field of light field reconstruction technology. The method includes: performing diffuser plate mesh generation and coordinate mapping between the physical coordinate system and the pixel coordinate system; calibrating the global transmittance of the diffuser plate; calibrating the global light intensity-to-grayscale conversion relationship; and using the calibration to convert the global grayscale distribution of the image acquired by the camera into a global light intensity distribution. The technical solution of this application can quickly reconstruct the spatial light intensity distribution of the light source, greatly improving measurement efficiency and accuracy.

Description

Technical Field

[0001] This invention relates to the field of light field reconstruction technology, specifically to a light field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration. Background Technology

[0002] In today's era of rapid technological advancement, light field measurement technology, as a key technology in the field of optics, is widely used in various fields such as road lighting inspection, industrial lighting optimization, and high-end display device calibration. However, traditional light field measurement methods face a significant bottleneck—they are highly dependent on manual operation, which is not only inefficient but also prone to introducing operational errors, severely restricting the accuracy and reliability of the measurement results.

[0003] Traditional methods typically require discretizing the target plane into multiple regions, with an operator manually measuring the light intensity value point-by-point within each region using a handheld photometer. This process is not only time-consuming and labor-intensive, but also prone to measurement deviations due to the inaccuracy of manual positioning and slight shaking of the handheld device. Furthermore, the granularity of region division and the number of sampling points directly limit the spatial resolution of the reconstructed light intensity distribution surface, making it difficult for the measurement results to accurately reflect subtle changes in the true light field. Summary of the Invention

[0004] To address the problems of low efficiency and easy introduction of operational errors in existing optical field measurement methods due to their heavy reliance on manual operation, which seriously restricts the accuracy and reliability of measurement results, this application proposes an optical field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration.

[0005] To achieve the above objectives, the technical solution adopted in this application is: a light field reconstruction method based on coaxial optical path integration and dual-domain dynamic calibration, the method comprising:

[0006] Perform grid division of the diffuser plate and coordinate mapping between the physical coordinate system and the pixel coordinate system; calibrate the global transmittance of the diffuser plate; calibrate the global light intensity-grayscale conversion relationship; and use the calibration to convert the global grayscale distribution of the image acquired by the camera into the global light intensity distribution.

[0007] Furthermore, the grid division of the diffuser plate includes: using a light shield of the same size as the diffuser plate, the light shield covers the entire diffuser plate, dividing the light shield into n × n regions of n rows and n columns, and making a cutout operation at the center of each region, the cutout size being a regular polygon tangent to the circular probe of the photometer used, where n is a natural number.

[0008] Furthermore, the coordinate mapping between the physical coordinate system and the pixel coordinate system includes:

[0009] Choosing the camera's pixel coordinate system as the reference; the length of the diffuser is 'a' and the width is 'b'; the origins of both the physical coordinate system and the pixel coordinate system on the diffuser are at the upper left corner of the coordinate system, with the x-axis pointing to the right and the y-axis pointing downwards as positive. Then, the center position of the i-th row and j-th column in the n×n region is represented by x in the physical coordinate system. i,j =(i-0.5)Δx,y i,j =(i-0.5)Δy, where Δx=a / n, Δy=b / n, and i and j are positive integers;

[0010] With the distance between the camera and the diffuser plate fixed at D, and the camera's auto-exposure function turned off, take an image of the diffuser plate. Then, through image processing, automatically identify the center coordinates (u) of the n×n cutout areas in the pixel coordinate system. i,j ,v i,j );

[0011] Connect the corresponding n×n pairs of points in the pixel coordinate system and the physical coordinate system (u i,j ,v i,j ) and (x i,j ,y i,j Using the least squares method to fit the mapping matrix H from the pixel coordinate system to the physical coordinate system, the following relationship exists:

[0012]

[0013] in, , , , , It is a real number.

[0014] Furthermore, calibrating the global transmittance of the diffuser plate includes: measuring the transmitted light intensity of n×n regions of the diffuser plate; measuring the true light intensity of n×n regions of the diffuser plate; and obtaining the global transmittance based on the transmitted light intensity and the true light intensity.

[0015] Furthermore, the measurement of the transmitted light intensity of the n×n regions of the diffuser plate includes: illuminating the reflective surface of the diffuser plate with a bar light source and transmitting the light through it; placing the probe of the photometer sequentially at positions corresponding to the transmission side of the diffuser plate and the n×n regions cut out by the light shield to measure the transmitted light intensity; turning off the bar light source to obtain the ambient light intensity of each n×n cutout region; and subtracting the ambient light intensity from the measured transmitted light intensity of each region sequentially to obtain the center (u) of each region in the camera's pixel coordinate system. i,j ,v i,j The actual transmitted light intensity I on the surface diffuser(i,j) Where (i,j) represents the region in row i and column j, and the whole is represented in matrix form. .

[0016] Furthermore, the measurement of the true light intensity of the n×n regions of the diffuser plate includes: removing the diffuser plate, leaving only the light shield; placing the photometer probe sequentially at the corresponding positions of the n×n regions cut out by the light shield to measure the true light intensity; turning off the bar light source to obtain the ambient light intensity of each n×n cutout region; and subtracting the ambient light intensity from the measured true light intensity of each region sequentially to obtain the center (u) of each region in the pixel coordinate system. i,j ,v i,j The actual true light intensity I on ) real(i,j) Where (i,j) represents the region in row i and column j, and the whole is represented in matrix form. .

[0017] Furthermore, obtaining the global transmittance based on the transmitted light intensity and the true light intensity includes: in the matrix and In the pixel coordinate system, each element at a corresponding position corresponds one-to-one. The ratio of the transmitted light intensity to the actual light intensity in different regions of the diffuser plate is the ratio of the transmitted light intensity to the actual light intensity in the n×n (u) pixels of the diffuser plate in the pixel coordinate system. i,j ,v i,j Local transmittance t at coordinates (i,j) , n×n t (i,j) A cubic spline fit is performed across the entire camera pixel coordinate system to obtain the distribution law of transmittance as a function of coordinates across the entire pixel coordinate system, which is represented as the global transmittance compensation matrix. The actual light intensity and the transmitted light intensity have the following conversion relationship on the diffuser plate:

[0018]

[0019] in, and The division is an element-by-element division.

[0020] Furthermore, the calibration of the global light intensity grayscale conversion relationship includes: keeping the diffuser plate position, bar light source position, light intensity, camera position, and camera exposure parameters unchanged; installing a replaceable neutral density filter assembly in front of the camera lens; calibrating the transmittance to be constant; and recording the transmittance of the neutral density filter as T when the light intensity entering the camera is within the sensor's linear response range. NDThe light shield is then placed back in front of the diffuser plate, and multiple photos of the diffuser plate's transmission side are taken continuously using a camera. Image processing is used to convert each photo into a grayscale Excel table in pixel coordinates, where each point represents the grayscale value of that pixel. The values ​​at the same location in multiple grayscale Excel tables are summed and averaged to obtain the average grayscale Excel table for that location of the diffuser plate. This average grayscale Excel table is then divided into n×n regions with the same shape and size as the cutout portion of the light shield, and the average grayscale G of each of these n×n regions is calculated. (i,j) Then turn off the light source, take a picture as the noise floor, and perform the same operation as above to obtain G. (ib,jb) ; each G (i,j) Subtract G from each (ib,jb) The final grayscale value G for each region is obtained. (i,j) As the center point (u) of each of the n×n regions i,j v i,j grayscale;

[0021] Based on each center point (u) in the n×n regions i,j v i,j Transmitted light intensity I at position ) diffuser(i,j) and the corresponding grayscale G (i,j) The ratio is the local transformation coefficient k at each center point location. (i,j) , n×n k (i,j) By performing cubic spline fitting across the entire camera pixel coordinate system, the distribution of the transformation coefficients across the entire pixel coordinate system as a function of coordinates is obtained, thus acquiring the spatially correlated light intensity-grayscale transformation matrix. The transmitted light intensity and the grayscale value on the image are given by the following formula:

[0022]

[0023] in, and Multiplication is element-wise multiplication.

[0024] Furthermore, the process of converting the global grayscale distribution of the image acquired by the camera into a global light intensity distribution using calibration includes: obtaining the calculation formula for the true light intensity in the pixel coordinate system by dual calibration of the global transmittance of the diffuser and the global light intensity grayscale conversion relationship, as shown in the following formula:

[0025]

[0026] in, For any point in the camera pixel coordinate system, the true light intensity of each pixel. Represented as Here, (i,j) represents the specific pixel coordinates in the pixel coordinate system;

[0027] To obtain the true light intensity distribution in the physical coordinate system of the diffuser, a coordinate mapping matrix H is used for transformation, and the calculation formula for the true light intensity distribution matrix in the physical coordinate system is as follows:

[0028]

[0029] in, Let be any point in the physical coordinate system of the diffuser.

[0030] On the other hand, this application also provides an optical field reconstruction system based on coaxial optical path integration and dual-domain dynamic calibration. The system includes a memory for storing computer program instructions and a processor for executing the program instructions. When the computer program instructions are executed by the processor, the system is triggered to execute the above-mentioned optical field reconstruction method based on coaxial optical path integration and dual-domain dynamic calibration.

[0031] The optical field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration proposed in this application have the following advantages compared with the prior art:

[0032] 1. Improved Measurement Efficiency: Existing technologies rely on manual point-by-point measurements using handheld photometers, which is time-consuming, labor-intensive, and prone to errors due to inaccurate positioning or equipment movement. This application employs automated single-shot imaging, integrating a coaxial optical path (strip light source, diffuser, and camera arranged along the same optical axis) to reconstruct the spatial light intensity distribution of the entire light source in a single shot, reducing measurement time from several hours to minutes. Combined with image processing and matrix operations, the global light intensity distribution can be calculated simultaneously, avoiding random errors from manual operation.

[0033] 2. Improved measurement accuracy: Traditional methods, with their manually divided grid density, limit spatial resolution and make it difficult to capture subtle changes in the light field (such as edge light intensity attenuation). Furthermore, the directionality of the light source causes a difference in the receiving angle between the camera and the photometer (cosine effect), leading to inaccurate calibration coefficients.

[0034] This application employs dual-domain dynamic calibration: Global transmittance compensation: By fitting the transmittance distribution of the diffuser plate with cubic splines, local overexposure / underexposure is eliminated, significantly improving the optimization of the dynamic calibration range. Light intensity-to-grayscale conversion matrix: Combined with a neutral density filter to adjust the light intensity, ensuring that the camera sensor operates within the linear range, significantly reducing the grayscale-to-light intensity conversion error.

[0035] 3. Comprehensive Expansion of Applicable Scenarios: Traditional methods are only suitable for static, small-scale light field measurements, and are powerless against dynamic or complex light fields (such as LED arrays and laser projections). This application can capture changes in the light field over time through real-time imaging and rapid calculation; it supports the overlay of thermal images from multiple light sources to achieve three-dimensional spatial light intensity distribution reconstruction.

[0036] 4. The dual-domain mapping method of "physical coordinate system - pixel coordinate system" and the dual calibration method of "transmittance-grayscale conversion" proposed in this application solves the two major problems of angle sensitivity and spatial non-uniformity in traditional calibration. Moreover, the use of cubic spline fitting instead of linear interpolation makes the distribution of transmittance and conversion coefficient smoother, significantly reduces interpolation error, and improves interpolation effect.

[0037] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0038] Figure 1 This is a flowchart of the optical field reconstruction method based on coaxial optical path integration and dual-domain dynamic calibration according to the present invention;

[0039] Figure 2 This is a schematic diagram of the system components after adding a light shield according to the present invention;

[0040] Figure 3 This is a mesh partitioning diagram in the physical coordinate system according to the present invention when n is 3;

[0041] Figure 4 This is a grid division diagram of the pixel coordinate system according to the present invention when n is 3;

[0042] Figure 5 This is a schematic diagram showing the relative positions of the two light shields and the diffuser plate according to the present invention;

[0043] Figure 6 This is a schematic diagram of the component placement when measuring real light intensity according to the present invention;

[0044] Figure 7 This is a schematic diagram of the relative positions of system components according to the present invention. Detailed Implementation

[0045] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings, so as to more clearly understand the purpose, features and advantages of this invention. It should be understood that the embodiments shown in the drawings are not intended to limit the scope of this invention, but are only for illustrating the essential spirit of the technical solutions of this invention. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0046] Unless the context requires otherwise, throughout the specification and claims, the word “comprising” and its variations, such as “including” and “having”, shall be understood to have an open, inclusive meaning, that is, to be interpreted as “including, but not limited to”.

[0047] Throughout this specification, references to "an embodiment" or "an embodiment" indicate that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Therefore, the appearance of "in an embodiment" or "an embodiment" in various places throughout the specification does not necessarily refer to the same embodiment. Furthermore, a particular feature, structure, or characteristic may be combined in any manner in one or more embodiments.

[0048] The singular forms “a” and “the” used in this specification and the appended claims include plural references unless otherwise expressly stated herein. It should be noted that the term “or” is generally used to mean “and / or” unless otherwise expressly stated herein.

[0049] In the following description, in order to clearly demonstrate the structure and working method of the present invention, a number of directional terms will be used. However, terms such as "front", "back", "left", "right", "outside", "inside", "outward", "inward", "up", and "down" should be understood as convenient terms and not as limiting terms.

[0050] The implementation details of the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The following content is only for the convenience of understanding the implementation details and is not necessary for implementing this solution.

[0051] To overcome the technical shortcomings of existing technologies, such as the inaccuracy of manual positioning and the slight shaking of handheld devices leading to deviations in measurement results and making it difficult for the measurement results to accurately reflect the subtle changes in the real light field, this invention proposes a light field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration. By combining system components such as a strip light source, a diffuser plate, and a camera along the same optical axis, the spatial light intensity distribution of the light source can be quickly reconstructed in a single shot, greatly improving measurement efficiency and accuracy.

[0052] It should be noted that the strip light source is a linear lighting device specifically designed for machine vision. Its core structure consists of high-density LED beads arranged in a straight line or arc, encapsulated in a long strip-shaped housing. Depending on the application requirements, it can be used individually or in combination (such as in a ring, coaxial, or cross arrangement) to form flexible lighting solutions. The diffuser plate is an optical element whose core function is to achieve uniform illumination by scattering light. Its working principle is based on the diffuse reflection characteristics of the surface microstructure: when light shines on the surface of the diffuser plate, due to the presence of tiny bumps or coating particles, the light is scattered in multiple directions rather than reflected in a single direction. This characteristic results in a uniform distribution of emitted light, avoiding strong light spots and shadows. Furthermore, the camera in this embodiment uses, but is not limited to, an industrial CCD camera. Industrial CCD cameras undergo precise calibration before leaving the factory, and their pixel response curves exhibit strictly linear characteristics. This means that there is a stable linear conversion relationship between the grayscale value of the image captured by the camera and the actual light intensity.

[0053] This implementation uses an industrial CCD camera. Since there is a stable linear conversion relationship between the grayscale values ​​of the image captured by the industrial CCD camera and the actual light intensity, the light intensity distribution of the original light field can be inferred by analyzing the grayscale distribution of the industrial CCD camera image. However, directly using an industrial CCD camera to photograph the light source and attempting to establish a correspondence between grayscale and light intensity will lead to two problems. First, the light source (such as LEDs or lasers) is directional, causing differences between the camera and the photometer due to different angles of light reception (cosine effect), which in turn makes it impossible to directly calibrate the relationship between grayscale and light intensity of the industrial CCD camera or results in inaccurate calibration coefficients. Second, the light source exhibits a hotspot effect (i.e., a non-uniform distribution with high light intensity at the center and attenuation at the edges). Direct imaging will result in insufficient dynamic range of the industrial CCD camera (overexposure at the center, underexposure at the edges, slow grayscale changes in the saturation region or logarithmic growth in the low-light region), leading to systematic errors in the calibration of grayscale and light intensity of the industrial CCD camera. To address the first issue, using a diffuser plate eliminates direction dependence: it converts directional light into Lambertian radiation (isotropic), making light intensity dependent only on spatial location and independent of the incident angle. This achieves calibration consistency, ensuring that industrial CCD camera and photometer measurements reflect only the spatial distribution of the true light field, eliminating angle sensitivity and establishing a linear mapping between position and light intensity. To address the second issue, using a diffuser plate improves spatial uniformity: it achieves spatial light intensity homogenization through multiple scattering, eliminating local overexposure / underexposure. Peak light intensity is reduced, preventing saturation in the industrial CCD camera, while preserving edge signal-to-noise ratio, thus optimizing the dynamic calibration range.

[0054] Furthermore, a diffuser plate can be used in conjunction with a neutral density filter. While a neutral density filter can linearly attenuate the overall light intensity, it cannot address spatial non-uniformity. By combining the two, the light intensity entering the industrial CCD camera can be flexibly adjusted while ensuring the spatial uniformity of the light source, ensuring that all light intensity entering the industrial CCD camera is within the linear range of the sensor.

[0055] The flowchart of the specific implementation method of light field reconstruction based on coaxial optical path integration and dual-domain dynamic calibration in this application embodiment is as follows: Figure 1 As shown. It should be noted that this application uses light intensity as a uniform metric. The method specifically includes the following steps:

[0056] S1. Perform diffuser mesh generation and coordinate mapping between the physical coordinate system and the pixel coordinate system.

[0057] Specifically, in a dark environment, a diffuser is placed at a fixed distance *d* from the strip light source. The aspect ratio of the diffuser must match that of the camera's field of view (FOV) to facilitate the subsequent conversion from the camera's pixel coordinate system to the diffuser's physical coordinate system. The FOV refers to the horizontal or vertical spatial range within which the camera can clearly image at a specific distance, usually expressed in angles (°) or actual dimensions (such as width and height). The diffuser is divided into n × n regions (e.g., 3 × 3 = 9 regions). Specifically, a light shield is fabricated using 3D printing technology. The light shield is the same size as the diffuser and can completely cover it. The light shield is then evenly divided into n × n regions, with a cutout at the center of each region. The cutout size is a regular polygon tangent to the circular probe of the photometer used, facilitating the measurement of the center position of each region. A detailed placement diagram is shown below. Figure 2 As shown.

[0058] Next, we need to calculate the center coordinates of each cutout area on the diffuser (or light shield; since they are in close contact during use, their coordinate systems can be considered the same; for ease of understanding, the following descriptions of coordinate systems will refer to them as diffuser). It's important to note that this involves choosing between the physical coordinate system on the diffuser and the camera's pixel coordinate system. If the physical coordinate system is chosen, decimals will appear when converting to pixel coordinates. Rounding will introduce errors with each rounding, leading to accumulated errors in subsequent calculations and fitting. Choosing the pixel coordinate system avoids this problem. Therefore, this application uses the pixel coordinate system as the basis for the following representations and conversions. If the length of the diffuser is *a* and the width is *b*; the origins of both the physical and pixel coordinate systems are at the upper left corner of the coordinate system, and the x-axis is positive to the right and the y-axis is positive downwards, then the center position of the i-th row and j-th column in the n×n regions can be represented as x in the diffuser's physical coordinate system. i,j=(i-0.5)Δx,y i,j =(i-0.5)Δy, (i,j=1,2,...,n), where Δx=diffuse plate length a / n, Δy=diffuse plate width b / n, that is, the center coordinates (x,j) of these n×n grid regions. i,j ,y i,j It is known that when n is 3, the mesh division in the physical coordinate system is as follows: Figure 3 As shown.

[0059] With the distance between the camera and the diffuser plate fixed at D, and the camera's automatic exposure function turned off, an image of the diffuser plate is captured. Through image processing (such as edge detection and contour extraction), the center coordinates (u) of the n×n hollowed-out areas in the camera's pixel coordinate system are automatically identified. i,j ,v i,j When n is 3, the grid division in the pixel coordinate system is as follows: Figure 4 As shown.

[0060] Connect the corresponding n×n pairs of points in the pixel coordinate system and the physical coordinate system ((u i,j ,v i,j ) and (x i,j ,y i,j Using the least squares method, we fit the mapping matrix H from pixel coordinates to physical coordinates, resulting in the following relationship:

[0061]

[0062] in, , , , , It is a real number.

[0063] In subsequent calibration and measurement, the pre-calculated physical coordinates are no longer relied upon. Instead, the corresponding physical coordinates are deduced by using the pixel coordinates of the image obtained by the camera and the mapping matrix H, thus avoiding the forced alignment error from physical coordinates to pixel coordinates.

[0064] S2, calibrate the global transmittance of the diffuser plate.

[0065] Because diffuser plates have a certain degree of spatial non-uniformity, there may be slight differences in transmittance in different areas, such as high transmittance in the center and low transmittance at the edges. However, manufacturers usually label a single fixed transmittance value. If the diffuser plate and its labeled fixed transmittance are used directly, it may lead to distortion of light intensity distribution, which may further cause errors in subsequent fitting.

[0066] For the reasons mentioned above, it is necessary to pre-calibrate the global transmittance of the diffuser plate, that is, to calibrate the global transmittance of the diffuser plate, in order to reduce the error caused by the spatial non-uniformity of the diffuser plate. Specifically, the transmittance is calculated, which is the ratio of the transmitted light intensity to the actual light intensity in different areas of the diffuser plate, and all the obtained ratios are fitted over the entire area to obtain the distribution law of the transmittance of the diffuser plate.

[0067] It is important to note that before calculating transmittance, the following three relevant calculation formulas are presented.

[0068] I. The conversion relationship between the true incident light intensity and the transmitted light intensity at a single point on the diffuser plate is as follows:

[0069]

[0070] Where t is the light intensity transmittance between the incident side and the transmitted side.

[0071] II. To clearly distinguish between standard matrix division and division of elements at the same position, this application defines a special operator:

[0072]

[0073] The element-wise division between matrices A and B is represented by the following rules:

[0074]

[0075] Here, A and B are identical matrices of arbitrary dimensions.

[0076] III. To clearly distinguish between standard matrix multiplication and element-wise multiplication, this application defines a special operator:

[0077]

[0078] The element-wise multiplication of matrices A and B is represented by the following formula:

[0079]

[0080] Here, A and B are identical matrices of arbitrary dimensions.

[0081] The specific steps are as follows:

[0082] First, measure the transmitted light intensity in n×n regions of the diffuser plate.

[0083] Specifically, the reflective surface of the diffuser is illuminated by the strip light source to be tested, and the transmitted light is transmitted through it. The probe of the photometer is placed sequentially at positions corresponding to the n×n areas cut out by the light shield on the transmission side of the diffuser to measure the transmitted light intensity. To ensure the accuracy of the measurement, five consecutive measurements are taken at the same position, and the average value is taken. Because the ambient light intensity may affect the measurement and calibration of the diffuser during the measurement process, the ambient light intensity needs to be measured and subtracted. The specific steps are as follows: turn off the strip light source, repeat the above measurement steps to obtain the ambient light intensity of each n×n cutout area, and subtract the ambient light intensity from the previously measured transmitted light intensity of each area to obtain the center (u) of each area in the pixel coordinate system. i,j ,v i,j The actual transmitted light intensity I on the surface diffuser(i,j) Where (i,j) represents the region in row i and column j, which can be represented as a matrix. If n=3, the matrix form is as shown in the following formula.

[0084]

[0085] To facilitate determining the measurement position of the photometer probe, an additional 3D-printed light shield of the same size and with the same cutout positions can be placed close to the transmission side of the diffuser plate. The photometer probe is then placed sequentially in each cutout area. Since the size of each cutout area is a square tangent to the circular probe of the photometer, the position of the photometer probe in each area is uniquely determined. The coordinates of the probe center and the previously calculated center of the cutout area (x, y, y) are then used to determine the position of the photometer probe. i,j ,y i,j ) or (u i,j ,v i,j Corresponding to the diagram. The relative positions of the two light shields and the diffuser are shown in the diagram. Figure 5 As shown.

[0086] Next, the actual light intensity of n×n regions of the diffuser plate is measured.

[0087] Specifically, remove the aforementioned diffuser, leaving only the light shield, such as... Figure 6 As shown. The photometer probe was placed sequentially at the corresponding positions of the n×n areas cut out of the light shield to measure the true light intensity. To ensure the accuracy of the measurement values, five consecutive measurements were taken at the same position, and the average value was taken. Because the ambient light intensity may affect the measurement calibration diffuser during the measurement process, the ambient light intensity needs to be measured and subtracted. The specific steps are as follows: turn off the bar light source, repeat the above measurement steps to obtain the ambient light intensity of each n×n cutout area, and subtract the ambient light intensity from the true light intensity of each area obtained previously to obtain the center (u) of each area in the pixel coordinate system. i,j ,v i,j The actual true light intensity I on ) real(i,j)Where (i,j) represents the region in row i and column j, which can be represented as a matrix. If n=3, the matrix form is as shown in the following formula.

[0088]

[0089] The two matrices above and In the pixel coordinate system, at each corresponding position (i.e., (u i,j ,v i,j The elements of the array correspond one-to-one, and their ratio is the local transmittance t of the transmission plate at that location. (i,j) This will yield n×n (u) pixels in the pixel coordinate system. i,j ,v i,j Local transmittance t at coordinates (i,j) , n×n t (i,j) By performing cubic spline fitting across the entire camera pixel coordinate system (or diffuser coordinate system), the distribution law of transmittance as a function of coordinates across the entire pixel coordinate system can be obtained, i.e., the global transmittance compensation matrix. The conversion relationship between actual light intensity and transmitted light intensity on the diffuser plate is shown in the following equation. The reason for using cubic spline fitting instead of linear interpolation fitting is that the fitted transmittance distribution of the diffuser plate is smoother, resulting in better interpolation performance. Here, "global area" refers to all pixels within the image range captured by the camera.

[0090]

[0091] in, and The division is the element-by-element division mentioned above, that is, dividing the elements at corresponding positions.

[0092] The above steps complete the global transmittance calibration of the diffuser used. Next, only the (u,v) coordinates in the camera pixel coordinate system need to be provided to calculate the global transmittance compensation matrix. The transmittance value at that point is retrieved, and the conversion between the actual light intensity and the transmitted light intensity is completed.

[0093] S3, calibrate the global light intensity grayscale conversion relationship.

[0094] Specifically, an industrial CCD camera with factory-calibrated optical settings was used. This camera exhibits good pixel response and linearity of incident light intensity without geometric distortion, and demonstrates a certain conversion relationship between light intensity and grayscale. Keeping parameters such as the diffuser plate position, light source position, light intensity, camera position, and camera exposure constant, a replaceable neutral density filter assembly was installed in front of the camera lens. Its calibrated transmittance was constant (dimension 1). By replacing neutral density filters with different transmittance models, a model was selected that ensured the light intensity entering the camera fell within the sensor's linear response range. The transmittance of this neutral density filter was recorded as T. ND Place the aforementioned light shield back in front of the diffuser. Now, only light from n×n areas can pass through the diffuser. Take five consecutive photographs of the diffuser's transmission side. The relative positions of the system components at this point are shown in the diagram below. Figure 7 As shown.

[0095] Image processing converts each captured image into a grayscale Excel table in pixel coordinates (i.e., each point's value is its grayscale value). The values ​​at the same location in each of the five grayscale Excel tables are summed and averaged to obtain the average grayscale Excel table for that location of the diffuser, reducing random errors during shooting. In image processing, a "grayscale Excel table" refers to converting each pixel of the image into a cell in a table, filling the cell with the pixel's grayscale value (an integer from 0 to 255), ultimately forming a two-dimensional table with pixel coordinates as row and column indices and grayscale values ​​as content. The average grayscale Excel table is divided into n×n regions of the same shape and size as the cutout portion of the light shield (these square regions are tangent to the circular probe of the diffuser), and the average grayscale value G of each of these n×n regions is calculated. (i,j) Then turn off the light source, take a picture as the noise floor, and perform the same operation as above to obtain G. (ib,jb) Each G (i,j) Subtract G from each (ib,jb) The final grayscale value G for each region is obtained. (i,j) As the center point (u) of each of the n×n regions i,j v i,j The reason for keeping the camera position constant is to avoid the influence of different shooting distances. The reason for taking the average gray level of each divided area instead of directly exporting the corresponding gray level value of this point in the average gray level Excel table is because the photometer measures the total diffuse light intensity within the probe area.

[0096] Based on each center point (u) in the n×n regions i,j v i,j Transmitted light intensity I at position ) diffuser(i,j)(Measured during the aforementioned diffuser calibration process) and the corresponding grayscale G (i,j) The ratio is the local transformation coefficient k at each center point location. (i,j) , n×n k (i,j) Performing cubic spline fitting across the entire camera pixel coordinate system (or diffuser coordinate system) yields the distribution of the transformation coefficients across the entire pixel coordinate system as the coordinates change; this is the spatially correlated light intensity-to-grayscale transformation matrix. The transmitted light intensity and the grayscale values ​​in the image have the following conversion relationship. Cubic spline fitting is used here instead of linear interpolation because the fitted conversion coefficients have a smoother distribution and the interpolation effect is better. Here, "global domain" refers to all pixels within the image range captured by the camera.

[0097]

[0098] in, and The multiplication is the element-wise multiplication described above, that is, multiplying corresponding elements. This is because the gray level G at the center point of each region is [value missing] during the calibration process. (i,j) Since the background noise has already been deducted, it can be assumed that the conversion formula does not contain an intercept term. In the subsequent actual conversion, the grayscale value obtained from the camera image also needs to be deducted in advance from the grayscale value of the actual background noise of the current location environment.

[0099] S4. Using the calibration relationship between S2 and S3, the global grayscale distribution of the image acquired by the camera is converted into the global light intensity distribution.

[0100] The above steps have completed the calibration of the global transformation relationship between transmitted light intensity and image grayscale. Next, only the corresponding (u,v) coordinates in the camera pixel coordinate system need to be provided to obtain the global light intensity-grayscale transformation matrix. Find the conversion coefficient at that point and complete the conversion between transmitted light intensity and image grayscale.

[0101] By dual-calibrating the global transmittance of the diffuser plate (S2) and the global conversion relationship between light intensity and grayscale (S3), the global dual calibration of the light source can be achieved. The final formula for calculating the true light intensity in the pixel coordinate system is as follows:

[0102]

[0103] Based on this formula, in practical applications, by obtaining the grayscale matrix G(u,v) in the global pixel coordinate system from an image taken by the camera at a distance D from the diffuser's transmission surface, the true magnitude of the light intensity at a distance d from the bar light source in the pixel coordinate system can be calculated. It is important to note that once global dual calibration is performed, the distance D between the camera and the diffuser cannot be changed, as changing D would alter the pixel coordinate system and thus the global dual calibration. However, the distance d between the bar light source and the diffuser can be changed.

[0104] The division and multiplication operations described here are the element-by-element division and multiplication operations mentioned earlier, T ND It is a constant. For any point in the camera pixel coordinate system, the true light intensity of each pixel. It can be represented as Here, (i,j) represents the specific pixel coordinates in the pixel coordinate system. To obtain the true light intensity distribution in the physical coordinate system of the diffuser, a coordinate mapping matrix H can be used for transformation. The final formula for calculating the true light intensity distribution matrix in the physical coordinate system is as follows:

[0105]

[0106] in, Let be any point in the physical coordinate system of the diffuser.

[0107] Regarding the impact of mesh generation accuracy on the accuracy of light field simulation calculations, it is important to note that the selection of the mesh density parameter n×n directly affects the calculation accuracy of the light field simulation results. In engineering practice, the optimal mesh can be determined through a dual criterion: First, a termination condition is set as a threshold standard where the mean absolute error (MAE) is less than 1%. When the deviation between the simulation results and the measured data meets this criterion, it can be determined that the accuracy requirement has been met. Second, a dynamic check is used using the error decay rate. When the error decay rate decreases significantly during the mesh generation iteration process (e.g., from quadratic convergence to single convergence), it indicates that the accuracy saturation region has been entered. The corresponding mesh density at this point is the optimal solution that balances efficiency and accuracy. This dual-criteria strategy ensures the engineering reliability of the calculation results while avoiding the waste of computational resources caused by over-refining the mesh.

[0108] The following is a practical example of using the above method: measuring the true distribution of light intensity of the light field generated by a bar collimated light source within a fixed distance range.

[0109] First, under this bar collimated light source, the global transmittance of the diffuse plate and the global conversion relationship between light intensity and grayscale are double-calibrated, and the coordinate mapping matrix from the physical coordinate system to the pixel coordinate system is calculated.

[0110] Secondly, actual measurements were taken.

[0111] Specifically, the diffuser is placed sequentially at n positions at distances d, d+1, d+2…d+i-1…d+n-1 from the collimated bar light source, where (d+i-1) represents the diffuser at position i. The following identical operations are performed at each position: the collimated bar light source illuminates the diffuser, a camera is placed at a distance D from the diffuser, and an overall image of the diffuser at each position is captured; the image is converted into an average grayscale Excel spreadsheet; by combining the average grayscale Excel spreadsheet with the relationship and coordinate mapping matrix obtained from double calibration, the true light intensity distribution matrix of the diffuser at position i in the physical coordinate system can be calculated. , where (x,y) is any point in the physical coordinate system of the diffuser plate.

[0112] Finally, using the mathematical analysis software MATLAB, the true light intensity distribution matrix at n locations can be interpolated and fitted into a three-dimensional solid light intensity thermal map of the entire domain from position d to d+n-1 in a single direction.

[0113] In addition, this embodiment can also superimpose the thermal images of multiple light sources from different directions to obtain the superposition distribution of multiple light sources in three-dimensional space.

[0114] On the other hand, this application also provides an optical field reconstruction system based on coaxial optical path integration and dual-domain dynamic calibration. The system includes a memory for storing computer program instructions and a processor for executing the program instructions. When the computer program instructions are executed by the processor, the system is triggered to execute the aforementioned optical field reconstruction method based on coaxial optical path integration and dual-domain dynamic calibration.

[0115] This application proposes a light field reconstruction method and system based on coaxial optical path integration and dual-domain dynamic calibration. Through the core innovation of coaxial optical path integration and dual-domain dynamic calibration, it achieves a leap from "inefficient manual" to "automatic and accurate" light field measurement, supporting dynamic light field and multi-source superposition analysis. This provides a revolutionary tool for fields such as road lighting inspection, industrial lighting optimization, and high-end display equipment calibration. Its engineering practicality (e.g., automated processes, optimal mesh determination) and theoretical innovation (dual-domain calibration theory) further ensure its technological leadership. Compared with existing technologies, it has significantly beneficial effects, enabling rapid reconstruction of the spatial light intensity distribution of the light source from a single image, greatly improving measurement efficiency and accuracy.

[0116] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the invention by those skilled in the art without departing from the spirit and essence of the invention. Such modifications or substitutions should all fall within the scope of the invention, or any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the invention should be covered within the protection scope of the invention. Therefore, the protection scope of the invention should be determined by the scope of the claims.

Claims

1. An optical field reconstruction method based on coaxial optical path integration and dual-domain dynamic calibration, characterized in that, The method comprises: Grid division of the diffuser plate and coordinate mapping between a physical coordinate system and a pixel coordinate system are performed; the global transmittance of the diffuser plate is calibrated; the global light intensity gray scale conversion relationship is calibrated; and the global gray scale distribution of a picture obtained by a camera is converted into a global light intensity distribution by using the calibration.

2. The method of claim 1, wherein, The grid division of the diffuser plate comprises: using a light shield cover of the same size as the diffuser plate, the light shield cover covering the entire diffuser plate, n×n regions of n rows and n columns are evenly divided on the light shield cover, and a hollowing operation is performed at the center of each region, the hollowing size being a regular polygon circumscribing a circular probe of a used photometer.

3. The method of claim 2, wherein, The coordinate mapping between the physical coordinate system and the pixel coordinate system comprises: The pixel coordinate system of the camera is selected as a reference; the length of the diffusion plate is a, and the width is b; the origins of the physical coordinate system and the pixel coordinate system on the diffusion plate are both at the upper left corner of the coordinate system, and the x-axis is positive to the right, and the y-axis is positive downward; the center position of the i-th row and j-th column in the n*n region is represented in the physical coordinate system as x i,j =(i 0.5)Δx,y i,j =(i 0.5)Δy, wherein Δx=a / n, Δy=b / n, i and j are positive integers; The distance between the fixed camera and the diffuser plate is D, the automatic exposure function of the camera is closed, a diffuser plate image is shot, and the center coordinates (u i,j ,v i,j ) of the n x n hollowed-out areas in the pixel coordinate system are automatically identified through image processing; The n*n group of point pairs (u i,j ,v i,j ) and (x i,j ,y i,j ) corresponding to the pixel coordinate system and the physical coordinate system are fitted using the least square method to obtain the mapping matrix H from the pixel coordinate system to the physical coordinate system, and there is the following relationship: wherein , , , , are real numbers.

4. The method of claim 3, wherein, The calibration of the global transmittance of the diffuser plate comprises: measuring the transmitted light intensity of the n×n regions of the diffuser plate; measuring the real light intensity of the n×n regions of the diffuser plate; and obtaining the global transmittance based on the transmitted light intensity and the real light intensity.

5. The method of claim 4, wherein, The method comprises: illuminating the reflective surface of the diffusing plate with the bar-shaped light source to be measured and transmitting the light, placing the probe of the luminometer in turn at positions corresponding to the n×n regions of the diffusing plate on the transmission side and the n×n hollow regions of the light shield to measure the transmission light intensity; turning off the bar-shaped light source to obtain the ambient light intensity of each n×n hollow region, and subtracting the ambient light intensity from the measured transmission light intensity of each region in turn to obtain the actual transmission light intensity I diffuser(i,j) of each region center (u i,j ,v i,j ) in the pixel coordinate system of the camera, wherein (i,j) represents the i-th row and j-th column region, and the whole is represented in matrix form .

6. The method of claim 5, wherein, The measuring the real light intensity of the n×n regions of the diffusion plate comprises: removing the diffusion plate and leaving only the light shield; sequentially placing the probe of the luminometer at the positions corresponding to the n×n regions of the light shield and measuring the real light intensity; turning off the bar light source to obtain the ambient light intensity of each n×n region; and subtracting the ambient light intensity from the measured real light intensity of each region to obtain the actual real light intensity I i,j (v i,j ) on the center of each region (u real(i,j) , j) in the pixel coordinate system, where (i, j) represents the i-th row and j-th column region, and the whole is represented in the form of a matrix .

7. The method of claim 6, wherein, The obtaining the global transmittance based on the transmitted light intensity and the real light intensity comprises: in the matrix and , the elements at each corresponding position in the pixel coordinate system correspond one by one, and the ratio of the transmitted light intensity and the real light intensity of different regions on the diffusion plate is the local transmittance t i,j ,v i,j ) of the n×n (u (i,j) ,v (i,j) ) coordinates of the diffusion plate in the pixel coordinate system. The distribution rule of the transmittance changing with the coordinates in the pixel coordinate system is obtained by performing cubic spline fitting on the n×n t The real light intensity and the transmitted light intensity have the following conversion relationship on the diffusion plate: wherein and The division is element-wise division.

8. The method of claim 7, wherein, The calibration global light intensity gray scale conversion relationship comprises: keeping the positions of the diffuser plate and the bar light source, the light intensity, the camera position, and the camera exposure parameters unchanged, installing a replaceable neutral filter assembly in front of the camera lens, calibrating the transmittance as a constant, recording the transmittance of the neutral filter as T when the light intensity entering the camera is in the linear response range of the sensor ND ; the light shield is placed again in front of the diffuser plate, a plurality of photos of the transmission side of the diffuser plate are continuously taken by the camera, each of the taken photos is converted into a gray scale Excel table in a pixel coordinate system by image processing, the value of each point is a pixel gray scale value, the values of each same position in the plurality of gray scale Excel tables are added and averaged to obtain an average gray scale Excel table of the diffuser plate for the position; the average gray scale Excel table is divided into n×n regions with the same shape and size as the hollow part of the light shield, the average gray scale G (i,j) of each region in the n×n regions is calculated (ib,jb) ; each G (i,j) is subtracted from G (ib,jb) to obtain the final gray scale value G (i,j) of each region, which is the gray scale of each center point (u i,j , v i,j ) in the n×n regions. According to the ratio of the transmission light intensity I diffuser(i,j) and the corresponding gray scale G (i,j) at each center point (u i,j , v i,j ) position in n×n regions, the local conversion coefficient k (i,j) at each center point position is obtained, and n×n k (i,j) is fitted by a cubic spline in the global camera pixel coordinate system to obtain the distribution rule of the conversion coefficient varying with the coordinates in the global pixel coordinate system, and the global space-related light intensity gray scale conversion matrix is obtained , and the transmission light intensity and the gray scale value on the image are shown in the following formula: wherein and the multiplication is element-wise multiplication.

9. The method of claim 8, wherein, The conversion of the global gray scale distribution of the picture obtained by the camera into the global light intensity distribution by using the calibration comprises: through the double calibration of the global transmittance of the diffuser plate and the global light intensity gray scale conversion relationship, the calculation formula of the real light intensity in the pixel coordinate system is as shown in the following formula: wherein, is the real light intensity of each pixel point in the camera pixel coordinate system is expressed as where (i, j) represents the specific pixel point coordinates in the pixel coordinate system. If the real light intensity distribution in the physical coordinate system of the diffuser plate is to be obtained, a coordinate mapping matrix H is used for conversion to obtain the real light intensity distribution matrix in the physical coordinate system, and the calculation formula is as follows: wherein, is the physical coordinate of any point in the diffuse plate.

10. An optical field reconstruction system based on coaxial optical path integration and dual-domain dynamic calibration, the system comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein, When the computer program instructions are executed by the processor, the system is triggered to perform the method of any one of claims 1 to 9.

Images