High signal-to-noise ratio spectrum temperature measuring device and method for three-dimensional thermal state object

Abstract

The invention relates to the technical field of three-dimensional shape reconstruction and temperature measurement, and discloses a high signal-to-noise ratio spectrum temperature measurement device and method for a three-dimensional thermal state object. Comprising an active light projection module, a spectrum filtering module, a light splitting module, an upper computer, an imaging module and a synchronous controller, the active light projection module is used for emitting projection light rays to a to-be-measured thermal-state object, and after incoherent light reflected by the surface of the to-be-measured thermal-state object is divided into multiple beams through the light splitting module, near-main-axis light rays are filtered through the spectrum filtering module to obtain light rays of different wave bands. The light rays of different wave bands are respectively transmitted to the upper computer after images are acquired by the imaging modules; the synchronous controller is connected with the active light projection module and the imaging modules and used for controlling synchronous work of the active light projection module and the imaging modules, and the upper computer is used for reconstructing a three-dimensional temperature field of the surface of the to-be-measured thermal state object according to images obtained by the imaging modules. According to the invention, three-dimensional temperature field reconstruction with high signal-to-noise ratio performance can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional topography reconstruction and temperature measurement, and particularly to a high signal-to-noise ratio spectral temperature measurement device and method for three-dimensional hot objects. Background Art

[0002] In the fields of industrial inspection and scientific research, accurate temperature measurement of hot objects is crucial. However, traditional contact temperature measurement techniques have significant limitations: they not only may damage the structure of hot objects, but also it is difficult to achieve rapid and continuous measurement of large areas. This limitation is particularly prominent in scenarios where monitoring of high-temperature, high-pressure, or rapidly changing environments is required.

[0003] In the prior art, an infrared thermal imager is often used to achieve non-contact temperature measurement. Although the infrared thermal imager can cover a large measurement range, it cannot provide detailed topography information of the object surface. This information is crucial for understanding the heat conduction characteristics of the object surface, identifying local hot spots, and evaluating the uniformity of materials. Moreover, the existing infrared thermal imagers have insufficient anti-interference ability against strong ambient light. That is, the prior art cannot meet the requirements for comprehensive analysis of the surface of hot objects.

[0004] Therefore, it is necessary to invent a high signal-to-noise ratio spectral temperature measurement system and method for three-dimensional hot objects to realize the measurement of the three-dimensional topography and temperature distribution of hot objects, achieve comprehensive analysis of the surface of hot objects, and solve the defects in the existing methods. Summary of the Invention

[0005] In order to achieve high signal-to-noise ratio and high-resolution monitoring of the surface temperature and topography of hot objects, the present invention proposes a high signal-to-noise ratio spectral temperature measurement device and method for three-dimensional hot objects, integrating a high signal-to-noise ratio imaging technique based on multi-wavelength cooperation and a three-dimensional temperature measurement technique of multi-channel signal fusion, to achieve non-contact, rapid, and accurate acquisition of the three-dimensional topography information and temperature information of the surface of hot objects, and further achieve accurate measurement of the characteristics of hot objects.

[0006] To solve the above technical problems, the technical solution adopted by the present invention is: a high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects, comprising: an active light projection module, a spectral filtering module, a beam splitting module, a host computer, an imaging module, and a synchronization controller;

[0007] The active light projection module is used to emit projection light to the hot object to be measured. The incoherent light reflected by the surface of the hot object to be measured is split into multiple beams by the beam splitting module, and then the near-principal axis light is filtered by the spectral filtering module to obtain light of different bands. The light of different bands is respectively collected by an imaging module to acquire images and sent to the host computer;

[0008] The synchronization controller is respectively connected to the active light projection module and each imaging module, and is used to control the synchronous operation of the active light projection module and each imaging module. The host computer is used to reconstruct the three-dimensional temperature field of the surface of the hot object to be measured according to the images obtained by each imaging module.

[0009] The beam splitting module includes at least one broadband depolarization beam splitting prism. The spectral filtering module includes at least two spectral filter films, and each spectral filter film corresponds to a different wavelength band. There are at least two imaging modules.

[0010] The specific method for the host computer to reconstruct the three-dimensional temperature field of the surface of the hot object to be measured according to the images obtained by each imaging module is as follows:

[0011] Calculate the gradient-sensitive weight, frequency-domain energy weight, and interpolation coefficient corresponding to each imaging module; determine the interpolation factor according to the interpolation coefficient of each imaging module;

[0012] Obtain the images to be reconstructed obtained by each imaging module;

[0013] Fuse the images obtained by each imaging module according to the interpolation factor;

[0014] Calculate the surface temperature of the hot object to be measured according to the image information at each wavelength, and reconstruct the three-dimensional temperature field of the surface of the hot object to be measured in combination with the fused image.

[0015] The calculation formulas for the gradient-sensitive weight, frequency-domain energy weight, and interpolation coefficient corresponding to each imaging module are:

[0016]

[0017] w(x,y) = β·w g (x,y) + (1 - β)·w f (x,y);

[0018] Among them, w g (x,y) and w f (x,y) respectively represent the gradient-sensitive weight and the frequency-domain energy weight, w(x,y) represents the interpolation coefficient, I(x,y) represents the image of the projection light collected by the imaging module, F(k,l) represents the Fourier frequency transform signal of I(x,y); (k,l) represents different frequency components in the frequency domain, σ g represents the adaptive parameter, HF represents the high-frequency band region, ALL represents all regions, β ∈ [0,1], and is a coefficient related to the noise level.

[0019] When there are two imaging modules, the calculation formula for the interpolation factor is:

[0020]

[0021] Among them, w a (x, y) and w b (x, y) respectively represent the interpolation coefficients corresponding to two imaging modules.

[0022] The formula for fusing the images obtained by each imaging module is:

[0023] I c (x, y) = I a (x, y) - αI b (x, y);

[0024] Among them, I c (x, y) is the fused image after interpolation processing.

[0025] The calculation formula for the surface temperature of the hot object to be measured is:

[0026]

[0027] In the formula: I a and I b respectively represent the image information at wavelengths λ1 and λ2, and T represents the temperature to be obtained; and respectively represent the emissivities at wavelengths λ1 and λ2; C2 represents the second Planck radiation constant.

[0028] The imaging module is a CMOS high-speed camera.

[0029] The accuracy of the synchronization signal output by the synchronization controller is less than 50 ns, and the synchronization signal controls the exposure time, frame rate, and capture time of the CMOS high-speed camera through the external trigger interface on the CMOS high-speed camera.

[0030] In addition, the present invention also provides a high signal-to-noise ratio spectral temperature measurement method for a three-dimensional hot object, which is implemented based on the described device and includes the following steps:

[0031] Enable the active light projection module to emit projection light to the calibration plate;

[0032] Collect images of different bands through each imaging module;

[0033] According to the images collected by the imaging module, calculate the interpolation coefficients of each imaging module; determine the interpolation factor according to the interpolation coefficients of each imaging module;

[0034] Obtain the images of different bands obtained by each imaging module;

[0035] According to the interpolation factor, fuse the images obtained by each imaging module;

[0036] Calculate the surface temperature of the hot object to be measured according to the image information at each wavelength, and reconstruct the three-dimensional temperature field on the surface of the hot object to be measured by combining the fused images.

[0037] The present invention has the following beneficial effects compared with the prior art:

[0038] The present invention provides a high signal-to-noise ratio spectral temperature measurement device and method for three-dimensional hot objects. The active light projection module is used to send projection light to the object surface. The reflected light from the object surface is separated into different bands through a beam splitting prism and a spectral filter. Then, multiple cameras are used to acquire images at different bands. Finally, the images at different bands are fused by an adaptive interpolation method. The spectral-depth information is reconstructed and displayed through the fused images, and the three-dimensional temperature field is reconstructed. The present invention reconstructs the three-dimensional temperature field in a non-contact manner, avoiding the influence of high temperature on the device. Moreover, the adaptive difference method is used to fuse the images at different bands, improving the safety, accuracy, and achieving high signal-to-noise ratio performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the device scheme of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0040] Figure 1 It is a schematic structural diagram of a high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects provided in Embodiment 1 of the present invention;

[0041] Figure 2 It is a schematic working flow diagram of a high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects provided in Embodiment 1 of the present invention;

[0042] Figure 3 It is the image effect after processing by different interpolation factors alpha;

[0043] In the figure: 1 - hot object to be measured; 2 - broadband depolarization beam splitting prism; 3 - spectral filtering module; 5 - imaging module; 7 - active light projection module; 8 - host computer; 9 - synchronization controller. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described clearly and completely below. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] Embodiment 1

[0046] As Figure 1 shown, Embodiment 1 of the present invention provides a high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects, which is characterized by including: an active light projection module 7, a spectral filtering module 3, a beam splitting module 2, a host computer 8, an imaging module 5, and a synchronization controller 9.

[0047] In this embodiment, the active light projection module 7 is used to emit projection light to the to-be-measured hot object 1. The incoherent light reflected by the surface of the to-be-measured hot object 1 is split into multiple beams by the beam splitting module 2, and then the near-principal axis light is filtered by the spectral filtering module 3 to obtain light of different bands. The light of different bands is respectively collected by an imaging module 5 to acquire images and then sent to the host computer 8.

[0048] The synchronization controller 9 is respectively connected to the active light projection module 7 and each imaging module 5, and is used to control the synchronous operation of the active light projection module 7 and each imaging module 5. The host computer 8 is used to reconstruct the three-dimensional temperature field on the surface of the to-be-measured hot object 1 according to the images obtained by each imaging module.

[0049] Specifically, in this embodiment, the beam splitting module includes a broadband depolarizing beam splitting prism. The spectral filtering module 3 includes two spectral filter films, and each spectral filter film corresponds to a different band. There are two imaging modules 5. The wavelength of the broadband depolarizing beam splitting prism covers the visible light range, the splitting intensity ratio is 1:1, and the split beams have no dispersion and polarization. The spectral filter film can filter natural light and other artificial interference light sources, and only allows the near-principal axis light projected by the active light projection module 7 onto the to-be-measured hot object to pass through.

[0050] Furthermore, in this embodiment, the spectroscopic module may include two broadband depolarizing spectroscopic prisms, through which the near-main-axis light in the projection light reflected by the thermal object to be measured emitted by the active light projection module 7 can be divided into three beams of light; the spectral filtering module 3 includes three spectral filters, each spectral filter corresponds to a different band, and after the three beams of light pass through a spectral filter respectively, the light of the required band can be obtained from the projection light reflected by the object, and then the imaging module 5 has three, each of which collects an image corresponding to the projection light of one band. In addition, in this embodiment, the spectroscopic module may include more than two broadband depolarizing spectroscopic prisms, and then the near-main-axis light in the projection light reflected by the thermal object to be measured emitted by the active light projection module 7 can be divided into more than three beams of light, and finally images of more than three bands are collected.

[0051] Specifically, the imaging module adopts a CMOS high-speed camera, and the synchronization signal accuracy output by the synchronization controller 9 is less than 50ns. The synchronization signal controls the exposure time, frame rate and capture time of the CMOS high-speed camera through the external trigger interface on the CMOS high-speed camera. The active light projection module 7 adopts a Digital Light Procession (DLP) digital light projector. The synchronization controller 9 uses FPGA to generate a synchronization signal to control multiple imaging modules to work synchronously. The imaging device realizes information transmission with the host computer 8 through the GigE communication protocol.

[0052] Figure 2 A schematic diagram of a high signal-to-noise ratio spectral temperature measurement method for a three-dimensional hot object provided by an embodiment of the present invention. The working principle of the embodiment of the present invention is described below.

[0053] Under the condition of strong ambient light interference, the optical imaging system Figure 1 The medium-wideband depolarizing beam splitter prism 2, the composite filter devices 3 and 4 obtain images of different wavelength channels of the strong interference scene, and the mathematical form of the interpolation processing is:

[0054] I c (x,y)=I a (x,y)-αI b (x,y); (1)

[0055] Where: I c (x, y) is the fused image after interpolation processing; I a (x,y) and I b (x, y) are the image information of different band channels; α is the interpolation factor; (x, y) represents the spatial position coordinates of the image.

[0056] The present invention realizes high signal-to-noise ratio imaging based on multi-wavelength collaboration, such as Figure 1As shown in the figure, the image components obtained by the imaging device consist of the target signal and the atmospheric scattered light noise:

[0057] I(x,y) = I T (x,y) + I B (x,y); (2)

[0058] In the formula: I(x,y) is the image obtained by the imaging device; I T (x,y) is the target signal component; I B (x,y) is the atmospheric scattered light noise.

[0059] According to the optical theory, the different band components obtained after the light is split by the broadband depolarizing beam splitter prism and filtered by the spectral filtering module are independent of each other. Based on equations (1) and (2), the processed image for strong environmental light interference can be expressed as:

[0060]

[0061] In the formula: I c T (x,y) and I c B (x,y) are the signals related to the target signal and the atmospheric scattered light noise in the interpolated processed image, and the expression forms are as follows:

[0062]

[0063] In the formula: and respectively represent the target information in the images of different band channels, and respectively represent the atmospheric scattered light noise in the images of different band channels. It can be seen from equations (3) to (5) that the quality of the interpolated processed image is closely related to the value of the interpolation factor α. When the value of the interpolation factor α is appropriate, the atmospheric scattered light noise can be completely filtered out theoretically, that is At this time, the image after the difference processing only includes the target information in the two band images, and the image quality is the best.

[0064] Such as Figure 3As shown, it is a schematic diagram of the image effect after data fusion using different interpolation factors. Each small image corresponds to a specific interpolation factor. As the interpolation factor alpha changes, the clarity and raster details of the image also change. When the interpolation coefficient is small (such as from 0.51 to 0.6), the image appears blurred and the details are not clear. As the interpolation coefficient increases, the image gradually becomes clear. Especially between 0.6 and 0.7, the edges and details of the image become more obvious. When the interpolation coefficient exceeds 1.0, the image begins to be distorted, which may be caused by over-interpolation. It can be seen from the figure that since a fixed interpolation factor is used for the entire image, the improvement of the signal-to-noise ratio of the fused image is limited. Therefore, the present invention uses an adaptive interpolation algorithm to find the optimal pixel-level interpolation factor α for effective noise elimination to achieve three-dimensional imaging with a high signal-to-noise ratio.

[0065] The adaptive interpolation algorithm adopted by the present invention dynamically adjusts the interpolation strategy according to the local features of the image (such as edges, textures, noise distribution, etc.). Different from the traditional fixed kernel function interpolation methods (such as bilinear and bicubic interpolation), the present invention analyzes the adaptive characteristics of the image content and selects the optimal interpolation weight or kernel function, so as to achieve a balance among denoising, detail preservation, and computational efficiency.

[0066] The projection light of the DLP digital light projector adopted in this embodiment is a light pattern with specific rules. Taking the stripe grating as an example, it has two characteristics: within a local window, the stripe gradient direction θ(x, y) follows the main direction distribution; its spatial frequency f changes slowly within the projection field of view. Therefore, the adaptive interpolation algorithm determines the interpolation factor α by setting double weights of gradient and frequency domain energy to achieve the idea of protecting the high-frequency region and smoothing the low-frequency region. The specific description is as follows.

[0067] First, the active light output by the DLP digital light projector is reflected by a calibration plate, and then after the imaging unit collects the images of each band channel, the calculation of the gradient-sensitive weight and the frequency domain energy weight is carried out.

[0068] The gradient-sensitive weight affects the processing of the high-frequency region. First, the Sobel operator is used to calculate the gradient amplitude of the pixel point:

[0069]

[0070] Among them, represents the gradient amplitude of the pixel point, I represents the image pixel value of the corresponding band channel, G x and G y represent the Sobel operator, and there are:

[0071]

[0072] Subsequently, an exponential decay function is introduced to suppress the interpolation smoothing in high-gradient regions. That is, the gradient-sensitive weight is expressed as:

[0073]

[0074] where σ g represents the adaptive parameter.

[0075] Then, the frequency-domain energy weight is calculated by performing local frequency-domain analysis on the window Ω through the fast Fourier transform. The calculation formula is:

[0076] F(k, l) = F{I(x, y)}; (9)

[0077] where F(k, l) represents the signal after Fourier transform, (k, l) represents different frequency components in the frequency domain, F{·} represents the Fourier transform, and I(x, y) represents the image pixel value of the corresponding band channel. Each (k, l) corresponds to a specific spatial frequency, k corresponds to the horizontal direction, and l corresponds to the vertical direction. As k and l increase, it corresponds to higher-frequency components such as edges and textures. By defining the region of the high-frequency band, the high-frequency energy ratio is calculated as the frequency-domain energy weight, that is:

[0078]

[0079] where HF represents the high-frequency band region and ALL represents all regions. Combining the gradient-sensitive weight and the frequency-domain energy weight, the interpolation coefficient is calculated:

[0080] w(x, y) = β · w g (x, y) + (1 - β) · w f (x, y); (11)

[0081] where β ∈ [0, 1] is a coefficient related to the noise level, which is dynamically adjusted:

[0082]

[0083] where σ0 is the noise threshold and γ is the adjustment factor.

[0084] When there are two imaging modules, the calculation formula for the interpolation factor is:

[0085]

[0086] where w a (x, y) and w b (x, y) respectively represent the interpolation coefficients corresponding to the two imaging modules.

[0087] When the number of band channels is greater than 2, the interpolation coefficients of each channel can be calculated, and then the interpolation factor can be calculated. The final image fusion formula is:

[0088] I c (x,y) = I a (x,y) - α1I b1 (x,y) - α2I b2 (x,y) …… - α n I bn (x,y); (14)

[0089] Among them, I a (x,y), I b1 (x,y), I b2 (x,y), …… I bn (x,y) represent the images of the 1st to n + 1th band channels respectively, and α1, α2, ……, α n represent the interpolation factors of the 2nd to n + 1th band channels. The calculation method is: calculate the interpolation coefficients of each band channel through the above process, and then divide them by the interpolation coefficient of the first band channel. The reciprocal is the interpolation factor of the corresponding channel. That is, the interpolation factors of each band channel can be calculated through the above formula (13), and finally the elimination of atmospheric scattered light noise is achieved.

[0090] The spectral temperature measurement algorithm based on radiation characteristics adopts the blackbody radiation law (also known as Planck's radiation law), which describes the relationship between the radiation rate of the electromagnetic energy radiated by an absolute blackbody and its frequency at a specific temperature, and can be expressed by the formula:

[0091]

[0092] In the formula: I(λ,T) represents the radiation energy of the blackbody, λ represents the wavelength of the radiation light, and T represents the temperature of the absolute blackbody; C1 and C2 represent the first and second Planck radiation constants respectively, and the values are C1 = 3.742×10 -16 W·m 2 , C2 = 1.4388×10 -2 W·m 2 . According to the above formula, the relationship between the absolute blackbody temperature, radiation wavelength and its radiation energy can be obtained. At a specific wavelength, the radiation energy of the blackbody is proportional to its temperature.

[0093] In actual temperature measurement, the hot object to be measured is a non-blackbody, and a new parameter needs to be introduced in the calculation of its radiation energy on the basis of formula (15):

[0094]

[0095] The newly introduced parameter ε in the formula represents the emissivity of the hot object to be measured, which characterizes the difference in radiation ability between the measured gray body and the absolute black body at the same temperature. The emissivity ε of the absolute black body is 1, and the emissivity ε of the gray body is < 1.

[0096] According to the above formula, the present invention can adopt two spectral temperature measurement methods:

[0097] (1) Using the adaptive interpolation algorithm to obtain the final wavelength λ0 of each unit pixel in the fused image. According to formula (16), it can be known that the temperature of the fused image can be obtained by using the monochromatic method, and the temperature can be calculated by the following formula:

[0098]

[0099] λ0 = λ1 - αλ2; (18)

[0100] where I c is the fused image after interpolation processing, ε represents the emissivity of the object to be measured, C1 and C2 respectively represent the first and second Planck radiation constants, λ1 and λ2 respectively represent the wavelengths of the two channels, and λ0 represents the fused wavelength. This temperature measurement method fuses the energy information at two wavelengths to solve the temperature, with higher accuracy. However, since the temperature of each unit pixel needs to be calculated separately according to the wavelengths obtained by different interpolation factors of each unit pixel, the calculation amount is relatively large.

[0101] (2) In addition, the present invention can also calculate the temperature by the two-color method. When the original images in two different wavelength channels are collected, the radiation energy ratio is obtained and then the temperature is calculated. Specifically, the two-color method temperature calculation formula is:

[0102]

[0103] In the formula: λ represents the wavelength of the radiation light, I a , I b respectively represent the image information at wavelengths λ1 and λ2, and T represents the temperature to be obtained; and respectively represent the emissivities at wavelengths λ1 and λ2; C2 represents the second Planck radiation constant, and the value is C2 = 1.4388×10 -2 W·m 2 . The calculation amount of this temperature measurement method is less than that of the first method.

[0104] Therefore, by combining the use of the adaptive interpolation method to perform pre-denoising on the image, high signal-to-noise ratio information is retained, and a high signal-to-noise ratio fused image after interpolation processing is obtained. Subsequently, based on formula (17) or (19), spectral temperature measurement is performed on the three-dimensional hot object, and the three-dimensional temperature field on the surface of the hot object 1 to be measured can be reconstructed with high signal-to-noise ratio.

[0105] Combined with the above principle, in this embodiment, the specific method for the host computer to reconstruct the three-dimensional temperature field on the surface of the hot object 1 to be measured based on the images obtained by each imaging module is as follows:

[0106] Calculate the gradient-sensitive weight, frequency-domain energy weight, and interpolation coefficient corresponding to each imaging module; determine the interpolation factor according to the interpolation coefficients of each imaging module;

[0107] Obtain the images to be reconstructed obtained by each imaging module;

[0108] Fuse the images obtained by each imaging module according to the interpolation factor;

[0109] Calculate the temperature at each position on the surface of the hot object to be measured according to the image information at each wavelength, and reconstruct the three-dimensional temperature field on the surface of the hot object (1) to be measured in combination with the fused image.

[0110] Among them, the calculation formulas for the gradient-sensitive weight, frequency-domain energy weight, and interpolation coefficient corresponding to each imaging module are formulas (8), (10), and (11) respectively.

[0111] Among them, the calculation formula for the interpolation factor is the above formula (13).

[0112] Among them, the formula for fusing the images obtained by each imaging module is formula (1).

[0113] Embodiment 2

[0114] Embodiment 2 of the present invention provides a high signal-to-noise ratio spectral temperature measurement method for a three-dimensional hot object, which is implemented based on the device described in Embodiment 1, and includes the following steps:

[0115] Make the active light projection module 7 emit projection light to the calibration plate;

[0116] Collect images in different bands through each imaging module;

[0117] Calculate the interpolation coefficients of each imaging module according to the images collected by the imaging module; determine the interpolation factor according to the interpolation coefficients of each imaging module;

[0118] Obtain the images obtained by each imaging module;

[0119] Fuse the images obtained by each imaging module according to the interpolation factor to obtain a three-dimensional image with high signal-to-noise ratio;

[0120] Calculate the surface temperature of the hot object to be measured according to the image information at each wavelength, and reconstruct the three-dimensional temperature field on the surface of the hot object 1 to be measured in combination with the fused image.

[0121] In summary, the present invention provides a high signal-to-noise ratio spectral temperature measurement device and method for three-dimensional hot objects. The active light projection module is used to send projection light to the object surface. The reflected light from the object surface is separated into different bands through a beam splitter prism and a spectral filter. Then, multiple cameras are used to acquire images in different bands. Finally, the images in different bands are fused by an adaptive interpolation method, and the spectral-depth information is reconstructed and displayed through the fused images to reconstruct the three-dimensional temperature field. The present invention reconstructs the three-dimensional temperature field in a non-contact manner, avoiding the influence of high temperature on the device. Moreover, the adaptive difference method is used to fuse the images in different bands, improving the safety, accuracy, and achieving high signal-to-noise ratio performance.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects, characterized in that, Including: An active light projection module (7), a spectral filtering module (3), a beam splitting module (2), a host computer (8), an imaging module (5), and a synchronization controller (9); The active light projection module (7) is configured to emit projection light to a hot object to be measured (1). The incoherent light reflected by the surface of the hot object to be measured (1) is split into multiple beams by the beam splitting module (2), and then the near-principal axis light is filtered by the spectral filtering module (3) to obtain light of different bands. The light of different bands is respectively collected by an imaging module (5) to acquire images and then sent to the host computer (8); The synchronization controller (9) is respectively connected to the active light projection module (7) and each imaging module (5), and is used to control the synchronous operation of the active light projection module (7) and each imaging module (5). The host computer (8) is used to reconstruct the three-dimensional temperature field on the surface of the hot object to be measured (1) according to the images obtained by each imaging module.

2. The high signal-to-noise ratio spectral temperature measurement device for a three-dimensional hot object according to claim 1, characterized in that, The beam splitting module includes at least one broadband depolarization beam splitting prism. The spectral filtering module (3) includes at least two spectral filter films, each spectral filter film corresponding to a different band, and there are at least two imaging modules (5).

3. The high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects according to claim 1, characterized in that, The specific method for the host computer to reconstruct the three-dimensional temperature field on the surface of the hot object to be measured (1) according to the images obtained by each imaging module is as follows: Calculate the gradient-sensitive weight, frequency-domain energy weight, and interpolation coefficient corresponding to each imaging module; determine the interpolation factor according to the interpolation coefficients of each imaging module; Obtain the images to be reconstructed obtained by each imaging module; Fuse the images obtained by each imaging module according to the interpolation factor; Calculate the surface temperature of the hot object to be measured according to the image information at each wavelength, and reconstruct the three-dimensional temperature field on the surface of the hot object to be measured (1) in combination with the fused image.

4. The high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects according to claim 3, characterized in that, The calculation formulas for the gradient-sensitive weight, frequency-domain energy weight, and interpolation coefficient corresponding to each imaging module are: w(x,y) = β·w g (x,y) + (1 - β)·w f (x,y); Among them, w g (x, y) and w f (x, y) represent the gradient-sensitive weight and the frequency-domain energy weight respectively, w(x, y) represents the interpolation coefficient, I(x, y) represents the image of the projection light collected by the imaging module, and F(k, l) represents the Fourier frequency transform signal of I(x, y); (k, l) represents different frequency components in the frequency domain, σ g represents the adaptive parameter, HF represents the high-frequency band region, ALL represents all regions, β ∈ [0, 1], and is a coefficient related to the noise level.

5. The high signal-to-noise ratio spectral temperature measurement device for a three-dimensional hot object according to claim 4, characterized in that When there are two imaging modules, the calculation formula for the interpolation factor is: Among them, w a (x, y) and w b (x, y) respectively represent the interpolation coefficients corresponding to two imaging modules.

6. The high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects according to claim 5, characterized in that The formula for fusing the images obtained by each imaging module is: I c (x, y) = I a (x, y) - αI b (x, y); Among them, I c (x, y) is the fused image after interpolation processing.

7. A high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects according to claim 5, characterized in that, The calculation formula for the surface temperature of the hot object to be measured is: where: I a and I b represent the image information at wavelengths λ1 and λ2 respectively, and T represents the temperature to be determined; and represent the emissivities at wavelengths λ1 and λ2 respectively; C2 represents the second Planck radiation constant.

8. The high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects according to claim 1, wherein The imaging module is a CMOS high-speed camera.

9. The high signal-to-noise ratio spectral temperature measurement device for three-dimensional hot objects according to claim 1, characterized in that, The accuracy of the synchronization signal output by the synchronization controller (9) is less than 50 ns, and the synchronization signal controls the exposure time, frame rate, and capture time of the CMOS high-speed camera through the external trigger interface on the CMOS high-speed camera.

10. A high signal-to-noise ratio spectral temperature measurement method for three-dimensional hot objects, implemented based on the device according to any one of claims 1 to 9, characterized in that, Including the following steps: Make the active light projection module (7) emit projection light to the calibration plate; Collect images of different bands through each imaging module; Calculate the interpolation coefficients of each imaging module according to the images collected by the imaging module; determine the interpolation factor according to the interpolation coefficients of each imaging module; Obtain the images of different bands obtained by each imaging module; Fuse the images obtained by each imaging module according to the interpolation factor; Calculate the surface temperature of the hot object to be measured according to the image information at each wavelength, and reconstruct the three-dimensional temperature field on the surface of the hot object to be measured (1) in combination with the fused image.

Images