A method for calculating non-visual effects of light environment based on spectral reconstruction of high dynamic range images

Abstract

The present invention proposes a method for calculating the non-visual effect of a light environment based on high dynamic range image spectrum reconstruction, comprising the following steps: step 1, in the environment to be tested, shooting and synthesizing a high dynamic range fisheye image; step 2, using the high dynamic range fisheye image obtained in step 1, performing spectrum reconstruction to obtain a relative spectrum; step 3, performing energy calibration on the relative spectrum to obtain an absolute spectrum; step 4, using the absolute spectrum to perform non-visual effect calculation to obtain a quantitative non-visual effect index for non-visual effect evaluation. The present invention has the characteristics of low cost and easy operation, and has good application prospects in the future research on the non-visual effect of lighting environments.

Description

Technical Field

[0001] The invention relates to a method for calculating non-visual effects of light environment, in particular to a method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction. Background Art

[0002] This section merely provides background information related to the present disclosure and is not necessarily prior art.

[0003] In modern life, indoors is the main office space for most people, so it is very important to carry out indoor light environment quality assessment.

[0004] With the improvement of light source technology and light source control level, lighting not only meets the basic visual functions of human eyes to see the size, color, shape, material and other basic visual functions of objects, but also has non-visual effects that affect the physical and mental state. People's attention to the quality of indoor lighting has shifted from the initial visual indicators such as whether it can see clearly and whether it will cause visual discomfort to the non-visual effects such as the impact of the lighting environment on biological rhythms, melatonin secretion, alertness, etc.

[0005] Typically, the non-visual effects of lighting need to be calculated by measuring the spectral distribution.

[0006] Photometric equipment for measuring spectra is expensive, and most are based on measurements at a single point or in a single direction, which makes it very difficult to obtain large-scale measurement data and fails to fully consider the impact of ambient reflection on the spectral distribution.

[0007] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0008] Purpose of the invention: The technical problem to be solved by the present invention is to provide a method for calculating the non-visual effects of light environment based on high dynamic range image spectral reconstruction in view of the shortcomings of the prior art.

[0009] In order to solve the above technical problems, the present invention discloses a method for calculating the non-visual effects of light environment based on high dynamic range image spectrum reconstruction, comprising the following steps:

[0010] Step 1, in the test environment, capture and synthesize a high dynamic range fisheye image;

[0011] Step 2, using the high dynamic range fisheye image obtained in step 1, performing spectral reconstruction to obtain a relative spectrum;

[0012] Step 3, performing energy calibration on the relative spectrum to obtain an absolute spectrum;

[0013] Step 4, using the absolute spectrum to calculate the non-visual effects, and obtain a quantitative non-visual effect index for non-visual effect evaluation.

[0014] Beneficial Effects

[0015] 1. The present invention reads the RGB three-channel values ​​and uses an integrating sphere device as a vignetting calibration tool, that is, uses the special properties of the integrating sphere to perform brightness compensation for high dynamic range fisheye images, thereby solving the brightness attenuation problem caused by the vignetting effect of the fisheye lens.

[0016] 2. The present invention analyzes the projection format of the fisheye lens, finds the correspondence between the pixel position under the distortion lens and the angle of the incident light, and analyzes the contribution of the illumination of each pixel of the high dynamic range fisheye image in the direction of the central axis of the fisheye lens, thereby realizing the calculation of the pixel brightness and RGB three-channel values ​​under the distortion lens.

[0017] 3. The present invention uses white light spots, i.e. The reflected spectrum of the light source Spectrum (λ), three-channel values ​​RGB, three stimulus values ​​XYZ, coordinates , the standard observer function , , The relationship between the RGB primary colors is calculated to obtain the reflectance spectrum of the three primary colors. , thus laying the foundation for relative spectrum calculation.

[0018] 4. The present invention combines the contribution value of the pixel RGB three-channel information under the distortion lens with the reflection spectrum of the three primary colors. The linear combination generates a relative spectrum, and the absolute spectrum is reconstructed through spectral energy calibration, thereby realizing the generation of the spectrum S from the pixel RGB three-channel values, solving the practical problem of high cost of spectral measurement equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more clear.

[0020] Figure 1 It is a system block diagram of the present invention.

[0021] Figure 2 It is a schematic diagram of the fisheye image vignetting calibration process.

[0022] Figure 3 It is a schematic diagram of the response curve of a standard observer at different wavelengths.

[0023] Figure 4It is a schematic diagram of the spectral basis curves of the three primary colors of red, green and blue calculated based on existing theories.

[0024] Figure 5 It is a numerical schematic diagram of EML calculated based on the reconstructed spectrum and the measured spectrum.

[0025] Figure 6 It is a numerical schematic diagram of CS calculated based on the reconstructed spectrum and the measured spectrum. DETAILED DESCRIPTION

[0026] The overall idea of ​​the present invention is as follows:

[0027] High dynamic range (HDR) images use 32-bit floating point numbers to store data. Each pixel records the color information of the three RGB channels. The brightness information of each pixel can be further calculated. The pixel value in the image is linearly related to the actual brightness value of the corresponding position in the lighting environment. After brightness calibration, it has great application potential in spatial lighting quality assessment.

[0028] The present invention proposes a method for recovering the corresponding relative spectral distribution from the RGB three-channel values ​​recorded in any given high dynamic range image, and after energy calibration, it is used for quantitative evaluation of non-visual effects of lighting.

[0029] The present invention reads the RGB three-channel values ​​of pixels from a high dynamic range image, calculates the relative reflectance spectral distribution in the high dynamic range image shooting scene in the form of linear combination based on the RGB three-primary color spectral distribution data under the D65 light source, and performs absolute energy calibration to obtain the absolute spectrum, which is applied to the quantification of non-visual effects of lighting.

[0030] The present invention is achieved through the following technical solutions:

[0031] 1. Spectral reconstruction based on high dynamic range images. This step uses pixel information recorded by high dynamic range images to reconstruct the relative reflectance spectrum distribution in the lighting scene, and further performs energy calibration to calculate lighting non-visual effect parameters. The method includes:

[0032] Low dynamic range (LDR) image capture and high dynamic range image synthesis; each high dynamic range image is synthesized from multiple low dynamic range images, using a fisheye lens camera fixed by a tripod and shooting a series of LDR images with different exposure times to ensure that the brightest point in the scene is not overexposed at the lowest exposure time, and the darkest point is not overexposed at the highest exposure time. The captured low dynamic range image sequence is synthesized into a high dynamic range fisheye image by high dynamic range synthesis software, and the synthesis software can be Photomatix pro software;

[0033] High dynamic range image pixel value reading; RGB three-channel values ​​of high dynamic range image pixels are read by software, and the software can use the hdrread function in MATLAB software;

[0034] 2. Generate a relative spectrum. This step specifically includes:

[0035] Fisheye lens brightness vignetting calibration; The vignetting effect of the fisheye lens causes the edge pixel information to decay, so vignetting calibration must be performed; The vignetting calibration method includes: placing the camera in an integrating sphere to shoot a series of low dynamic range images and using software to synthesize high dynamic range images, where the integrating sphere is a sphere of special material with a uniform diffuse reflection on the inner surface. In theory, the brightness values ​​of each pixel in the high dynamic range image synthesized after the low dynamic range images shot in the integrating sphere are equal. Based on this characteristic of the integrating sphere, it can be used for fisheye lens vignetting calibration. By selecting n pixels equidistantly from the center to the edge, according to the formula Calculate the brightness value of each pixel. Where i represents the i-th pixel, represents the brightness of the i-th pixel of the high dynamic range fisheye image, Represents the red channel value of the i-th pixel of the high dynamic range fisheye image, Represents the green channel value of the i-th pixel of the high dynamic range fisheye image, Represents the blue channel value of the i-th pixel of the high dynamic range fisheye image.

[0036] The ratio of the pixel brightness of each selected pixel point to the center brightness is used as the brightness attenuation factor, and the brightness attenuation curve f(d) is fitted, where d represents the number of pixels between the selected pixel point and the center pixel, and the reciprocal is taken to generate the brightness compensation curve 1 / f(d) about the distance from the edge pixel to the center pixel, that is, the brightness compensation factor is the reciprocal of the attenuation factor. When reading the brightness of each pixel, it is multiplied by the brightness compensation factor to complete the fisheye lens brightness vignetting calibration.

[0037] Calculation of RGB three-primary color spectrum under D65 environment. According to the definition of the International Illumination Commission, any measured reflectance spectrum Spectrum (λ) and the standard observer function specified by CIE The inner product can calculate its CIE XYZ coordinates, that is, the tristimulus value XYZ, as shown in formula 1, where λ represents the wavelength, It is the standard observer function, which expresses the human response to each wavelength of light in the form of a mathematical function and is used for color calculation; Indicates the spectral distribution of the D65 light source, and Spectrum(λ) represents an arbitrary reflection spectrum. In addition, the three primary colors RGB values ​​can be converted from the three stimulus values ​​XYZ through an M matrix, as shown in Formula 2, where the M matrix is ​​a 3*3 linear transformation matrix corresponding to sRGB coordinates to XYZ coordinates, which is defined according to the white light point coordinates YD65 and BT.709 standards, and the specific values ​​are shown in Formula 3.

[0038]

[0039] Therefore, under the D65 environment, the relationship between the three primary colors R, G, B and the spectral reflectance function Spectrum(λ) is expressed as shown in formula 4, where represents the coordinates of the white light point under the D65 light source environment, and ⨀ represents the Hadamard product, that is, the elements in the same position of the two matrices are multiplied to generate a new matrix.

[0040]

[0041] The solution of Spectrum(λ) can be calculated by using the modified optimization equation of Formula 4, namely Formula 5.

[0042]

[0043] By substituting the RGB color coordinates of (1,0,0), (0,1,0), (0,0,1) into Formula 5 to solve this optimization equation, the reflectance spectrum of the RGB primary colors under this D65 light source environment can be solved. .

[0044] The weighted average of the angle information of each channel of sRGB is solved. The sRGB color space is the most common color representation used in computer graphics. The white point of the sRGB space 〈1, 1, 1〉 corresponds to the CIE standard illuminant D65, which is a standardized spectral power distribution designed to roughly represent daylight. The fisheye image records the brightness and distribution information of pixels on a hemisphere in the shooting direction. Even after vignetting calibration, pixels at different azimuths contribute differently to the illumination in the shooting direction of the lens. Different fisheye lenses have different projection formats, generally including equidistant projection format, equi-stereoscopic projection format, orthogonal projection format, etc. The angle between the pixel Pi and the shooting direction on the high dynamic range fisheye image is calculated based on the lens projection format. , then the angle-weighted brightness of the pixel , calculate the weighted average brightness of the angle information of all pixels in this channel, and use the same method to calculate the weighted average of the angle information of the three channels, which are recorded as r, g, and b respectively.

[0045] High dynamic range fisheye image relative spectrum generation. The weighted average of the sRGB three-primary color angle information calculated by the above method, r, g, b and the sRGB three-primary color spectral reflectance basis obtained above are linearly combined to obtain the relative spectral distribution, that is, .

[0046] 3. Absolute spectrum calibration, this step includes:

[0047] Spectral energy calibration. The spectrum restored by the above method is a relative spectrum and cannot be directly used for the calculation of lighting parameters. The spectral energy calibration method includes: calculating the photopic index, i.e., illuminance, from the relative spectrum , the calculation formula is , and use an illuminance meter to measure the illuminance Ev in the direction of the fisheye lens shooting. The ratio of Ev to Ev' is used as the energy factor ef, and the absolute spectrum Relative spectrum The product of the energy factor ef is .

[0048] Finally, the calibrated absolute spectrum is obtained. , used for subsequent calculation and evaluation of non-visual effects.

[0049] 4. Calculate and evaluate non-visual effects, i.e., use the rhythm effect index quantification method based on the reconstructed spectrum to calibrate the absolute spectrum after calibration. Perform non-visual effect calculation and evaluation. The method comprises:

[0050] Quantification method of equivalent melanin lux (EML). The spectral sensitivity curve of ipRGC cells published by the CIE lighting official website is integrated with the above-recovered absolute spectrum and multiplied by the scaling factor to obtain the EML value. The calculation formula is: .in is the reconstructed absolute spectrum, is the spectral sensitivity curve of intrinsically photosensitive retinal ganglion cells.

[0051] Quantification method of circadian stimulation (CS). The spectral sensitivity curves of cones and rods published on the CIE lighting official website are used for corresponding calculations, as shown in formula 6. Among them, Indicates day and night light, is the spectral sensitivity curve of intrinsically photosensitive retinal ganglion cells, is the reconstructed absolute spectrum, represents the spectral sensitivity curve of S cone cells, represents the photopic luminous efficiency function, represents the luminous efficiency function of dark vision, Represents macular transmittance.

[0052]

[0053] The final quantitative values ​​of equivalent melanin lux (EML) and circadian stimulus (CS) were used as the basis for evaluating non-visual effects. Example

[0054] like Figure 1 As shown, the present invention is a spectral reconstruction based on high dynamic range images and a rhythmic effect calculation method based on the reconstructed spectrum. The method includes five steps: low dynamic range fisheye image acquisition, high dynamic range (HDR) fisheye image production, relative spectrum generation, absolute spectrum calibration, and non-visual effect calculation.

[0055] The three scenes selected are office, conference room and computer room as the scene examples of the present invention.

[0056] The low dynamic range fisheye image acquisition steps are carried out in the selected scene. The equipment involved are Canon 5DMark III camera, Sigma 8mm f3.5EX DG fisheye lens, tripod, Minolta CL-200A colorimeter. In the selected experimental scene, the camera equipped with a fisheye lens is mounted on a tripod and placed in the center of the experimental scene. It is connected to the computer with a data cable, and the computer controls the image shooting. In order to ensure that the brightest point in the scene is not overexposed at the lowest exposure time, and the darkest point is not overexposed at the highest exposure time, the first low dynamic range (LDR) image is taken with an exposure time of 1 / 8000s, and a group of LDR images are taken at equal intervals with different exposure times.

[0057] After the low dynamic range fisheye image was taken, a Minolta CL-200A color illuminance meter was fixed at the camera position to record the illuminance data at that position.

[0058] High dynamic range (HDR) image synthesis is completed by Photomatix pro6.1 software. A set of RAW format images of low dynamic range images taken are imported into Photomatix pro software and directly synthesized into an HDR image.

[0059] Image vignetting calibration is to compensate for the brightness attenuation of the image edge pixels. The specific steps are as follows: Figure 2As shown. The integrating sphere used in the example is a spherical device with a diameter of 2m. After multiple reflections on the inner surface of the integrating sphere, the brightness of each part of the sphere is the same. An LED light source is placed in the sphere, and a camera equipped with a Sigma 8mm f3.5EX DG fisheye lens is placed in the center of the integrating sphere to take a set of low dynamic range fisheye images, and the high dynamic range fisheye images are synthesized by Photomatix pro software. By selecting n points equidistantly from the center to the edge of the image, MATLAB software is used to calculate the high dynamic range fisheye image. Read the pixel brightness Y of the n points, and calculate the ratio of its brightness to the brightness of the center pixel as the brightness attenuation factor, fit the brightness attenuation curve f(d), where d represents the number of pixels between the selected pixel and the center pixel, and take the reciprocal to generate the brightness compensation curve 1 / f(d) about the distance from the edge pixel to the center pixel, that is, the brightness compensation factor is the reciprocal of the attenuation factor. When reading the brightness of each pixel, multiply it by the brightness compensation factor to complete the fisheye lens brightness vignetting calibration. The brightness compensation curve equation of this example lens is , x is the number of pixels between the calibrated pixel and the center pixel.

[0060] The brightness compensation equation obtained in the above steps is also applicable to the compensation of color information attenuation. The RGB color information of all pixels can be calibrated by this equation.

[0061] Since the above RGB three-channel values ​​do not incorporate angle information, that is, the problem of distortion caused by the fisheye lens is not taken into account, the incident angle θ of the light needs to be considered.

[0062] The fisheye image records the pixel brightness and angle distribution information on the hemisphere of the shooting direction. Even after vignetting calibration, pixels at different azimuths contribute differently to the illumination in the shooting direction of the lens. The angle between the pixel Pi and the shooting direction on the high dynamic range fisheye image is calculated based on the lens projection format. , the projection format of the example fisheye lens is equal solid angle, then the angle-weighted brightness of each pixel is , calculate the weighted average brightness of the angle information of all pixels in this channel, and use the same method to calculate the weighted average of the angle information of the three channels, which are recorded as r, g, and b respectively.

[0063] The relative spectrum reconstruction theory is developed based on the colorimetry and photometry foundations specified by the International Standard Illumination Commission. It is known that for any reflection spectrum Spectrum(λ) under the D65 environment, it is integrated with the standard observer function The CIE XYZ coordinates can be calculated as shown in formula 7, and the standard observer function image is as follows: Figure 3 As shown:

[0064]

[0065] For the sRGB space, which is the most common form of image representation, the calculation of the three primary colors RGB is as shown in Formula 8, where M is the conversion matrix from RGB to XYZ, which is a constant matrix defined based on the white light point coordinates and the BT.709 standard:

[0066]

[0067] Formula 8 and Formula 9 are rearranged to obtain Formula 10:

[0068]

[0069] For the solution of Spectrum(λ), it can be calculated by the modified optimization equation of Formula 10, that is, Formula 11:

[0070]

[0071] By substituting the RGB color coordinates of (1,0,0), (0,1,0), (0,0,1) into formula 11 to solve this optimization equation, the reflectance spectrum of the RGB three primary colors under this D65 light source environment can be solved. The corresponding base spectrum is plotted as Figure 4 shown.

[0072] The three sets of spectral bases calculated in the above steps and the three-channel values ​​r, g, b are linearly combined to reconstruct the relative spectrum. The relative spectrum reconstruction formula is: .

[0073] Spectral energy calibration refers to multiplying the relative spectrum by the energy coefficient to obtain the absolute spectral distribution, which is used to calculate all lighting indicators. On the one hand, the illuminance value is calculated by integrating the relative spectrum generated by the above steps with the photopic sensitivity function, that is, On the other hand, the illuminance value is measured by a CL-200A color illuminance meter placed in the same position as the camera. , the energy coefficient ef is the ratio of the measured value to the calculated value. That is, the absolute spectrum .

[0074] The present invention calculates the rhythmic effect of lighting by taking equivalent melanin lux EML and circadian rhythm stimulation CS as examples, and verifies its accuracy by comparing the rhythmic effect calculated from the actual spectrum measured by a CL-500A illuminometer.

[0075] The calculation formula of equivalent melanin lux EML is: Wherein S(λ) is the reconstructed absolute spectrum, and C(λ) is the spectral sensitivity curve of the intrinsically photosensitive retinal ganglion cells.

[0076] The EML value calculated by the present invention according to this formula after image acquisition, vignetting calibration, and energy calibration in the example scenario is compared with the EML value calculated by the spectrum measured by the CL-500A illuminance meter. Figure 5 As shown, the EML error between the spectrum restored by this method and the spectrum calculated by the CL-500A spectroradiometer is within 10%.

[0077] The calculation of equivalent melanopic lux CS is as follows: represents day and night light, C(λ) is the spectral sensitivity curve of intrinsic photosensitive retinal ganglion cells, S(λ) is the reconstructed absolute spectrum, represents the spectral sensitivity curve of S cone cells, represents the photopic luminous efficiency function, represents the luminous efficiency function of dark vision, Represents macular transmittance.

[0078]

[0079] The CS value calculated by the present invention according to this formula in the example scenario is compared with the CS value calculated by the spectrum measured by the CL-500A illuminance meter, as shown in FIG. Figure 6 As shown, the CS error between the spectrum restored by this method and the spectrum calculated by the CL-500A spectral radiometer is within 10%.

[0080] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, the invention content of the method for calculating the non-visual effects of the light environment based on high dynamic range image spectrum reconstruction provided by the present invention and some or all of the steps in each embodiment can be run. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0081] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on such understanding, the technical solutions in the embodiments of the present invention can be essentially or partly contributed to the prior art in the form of computer programs, i.e., software products, which can be stored in a storage medium and include several instructions for enabling a device including a data processing unit (which can be a personal computer, a server, a single-chip microcomputer, an MCU or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0082] The present invention provides a method and idea for calculating the non-visual effects of light environment based on high dynamic range image spectrum reconstruction. There are many methods and approaches to implement the technical solution. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention. All components not specified in this embodiment can be implemented by existing technologies.

Claims

1. A method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction, characterized in that: The following steps are involved: Step 1, in the test environment, capture and synthesize a high dynamic range fisheye image; Step 2, using the high dynamic range fisheye image obtained in step 1, performing spectral reconstruction to obtain a relative spectrum; Step 3, performing energy calibration on the relative spectrum to obtain an absolute spectrum; Step 4, using the absolute spectrum to calculate the non-visual effect and obtain a quantitative non-visual effect index for non-visual effect evaluation; The spectral reconstruction described in step 2 includes: Step 2-1, reading the RGB three-channel value of each pixel of the high dynamic range fisheye image, calculating the brightness value of each pixel, and performing vignetting calibration on the high dynamic range fisheye image; Step 2-2, calculating the relative spectrum of the high dynamic range fisheye image after vignetting calibration, specifically includes: Step 2-2-1, integrating the RGB three-channel values ​​of all pixels of the high dynamic range fisheye image after vignetting calibration into the angle information, and obtaining three-channel values ​​integrated with the angle information, which are recorded as: r, g and b, where the three channels are red, green and blue channels; Step 2-2-2, calculate three sets of spectral bases according to the white light point in sRGB space, denoted as: S r , S g and S b , i.e., the red, green, and blue channel spectral basis; Step 2-2-3, linearly combining the three-channel values ​​incorporating the angle information and the three groups of spectral bases, as shown below: S′=r·S r +g·S g +b·S b Among them, S′ is the result of spectral reconstruction, that is, the relative spectrum.

2. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 1, characterized in that: Capture and synthesize a high dynamic range fisheye image as described in step 1, including: Step 1-1, use a camera with a fisheye lens to take a group of low dynamic range fisheye images at equal intervals with different exposure times in the test environment, where the brightest point is not overexposed at the shortest exposure time, and the darkest point is not overexposed at the longest exposure time; Step 1-2, synthesize a set of low dynamic range fisheye images into a high dynamic range fisheye image.

3. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 2, characterized in that: The step 2-1 of performing vignetting calibration on the high dynamic range fisheye image comprises: Step 2-1-1, calculating the pixel brightness value of each pixel; Step 2-1-2, taking the ratio of the pixel brightness value of each pixel to the center brightness as the brightness attenuation factor, fitting the brightness attenuation curve f(d), where d represents the number of pixels between the selected pixel and the center pixel; Step 2-1-3, taking the inverse of the brightness attenuation curve, generating a brightness compensation curve 1 / f(d) as a brightness compensation factor; Step 2-1-4, read the brightness of each pixel and multiply it by the brightness compensation factor to complete the vignetting calibration.

4. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 3, characterized in that: The calculation of the pixel brightness value of each pixel described in step 2-1-1 includes: <h2 style=";text-align:left;direction:ltr">Y<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> =179×(0.2126R<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> +0.7152G<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> +0.0722B<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> ) Among them, i represents the i-th pixel, Y i represents the brightness of the i-th pixel of the high dynamic range fisheye image, R i Represents the red channel value of the i-th pixel of the high dynamic range fisheye image, G i Represents the green channel value of the i-th pixel of the high dynamic range fisheye image, B i Represents the blue channel value of the i-th pixel of the high dynamic range fisheye image.

5. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 4, characterized in that: The integration angle information described in step 2-2-1 includes: r i =R i cos(θ i ) g i =G i cos(θ i ) b i =B i cos(θ i ) Among them, r i Indicates the red channel value of the i-th pixel that incorporates the angle information, g i Indicates the green channel value of the i-th pixel that incorporates the angle information, b i Represents the blue channel value of the i-th pixel that incorporates the angle information, θ i represents the incident angle of light at the i-th pixel; The average of the three-channel values ​​of all pixels that have been integrated with the angle information is expressed as follows: Among them, r, g and b are three-channel values ​​that incorporate angle information.

6. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 5, characterized in that: The calculation of three sets of spectral bases described in step 2-2-2 includes: The constraint is: 0≤Spectrum(λ)≤1 Among them, D 65 represents the spectral distribution of the D65 light source, Spectrum(λ) represents an arbitrary reflection spectrum, and is the standard observer function, which is used to calculate the color of light with wavelength λ; ⊙ represents the Hadamard product, that is, the elements in the same position of two matrices are multiplied to generate a new matrix; T represents matrix transposition, Y D6 Represents the coordinates of the white light point in the sRGB space, M is a constant matrix, and R, G, and B are the values ​​of the three RGB channels in the sRGB space; When the RGB three-channel values ​​in the sRGB space are (1,0,0), (0,1,0) and (0,0,1), the obtained Spectrum (λ) is recorded as the red channel spectral basis S r , Green channel spectrum basis S g and the blue channel spectral basis S b .

7. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 6, characterized in that: The energy calibration of the relative spectrum in step 3 includes: Step 3-1, in the test environment described in step 1, use an illuminance meter to measure the illuminance in the shooting direction z, recorded as measured illuminance Ev; Step 3-2, calculating the illumination in the shooting direction of the high dynamic range fisheye image according to the relative spectrum, recorded as calculated illumination Ev′, and the calculation method is as follows: Ev′=683∫S′(λ)V(λ)dλ Among them, S′(λ) is the relative spectrum, V(λ) is the visual efficiency function of human eyes; Step 3-3, calculating the ratio of the measured illuminance Ev to the calculated illuminance Ev′ to obtain an energy calibration factor ef; Step 3-4, multiplying the relative spectrum by the energy calibration factor to obtain the absolute spectrum s in the shooting direction of the fisheye lens, which is expressed as follows: S(λ)=S(λ)′·ef Complete energy calibration.

8. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 7, characterized in that: Perform non-visual effect calculations as described in step 4, including: The quantitative calculation of equivalent melanin lux EML is performed as follows: EML=72983.25∫S(λ)C(λ)dλ Wherein, S(λ) is the absolute spectrum, and C(λ) is the spectral sensitivity curve of the intrinsically photosensitive retinal ganglion cells.

9. The method for calculating non-visual effects of light environment based on high dynamic range image spectrum reconstruction according to claim 8, characterized in that: The non-visual effect calculation described in step 4 also includes: The quantitative calculation of the circadian stimulation CS is performed as follows: Among them, CL A represents day and night light, C(λ) is the spectral sensitivity curve of intrinsic photosensitive retinal ganglion cells, S(λ) is the absolute spectrum, S cone represents the spectral sensitivity curve of S cone cells, V λ represents the photopic luminous efficiency function, V′ λ represents the luminous efficiency function of dark vision, mp λ Represents macular transmittance.

Images