Temperature measurement methods and modules, infrared thermometers, equipment and storage media

Abstract

This application provides a temperature measurement method and module, an infrared thermometer, device, and storage medium. The method acquires a reference signal and a target signal during the on and off phases, respectively, performs differential processing on them, extracts fluctuation information from the differential signals, and uses this fluctuation information as input to a noise prediction model for adaptive modeling and prediction. Through online updates of the weight vector, the predicted noise gradually approximates the actual target noise, and the target noise is then canceled out from the target signal during the off phase. In this application, the reference and target optical signals use the same optical channel in the epitaxial growth equipment, enabling the noise prediction model to accurately model and predict actual noise. This significantly improves the signal-to-noise ratio and effective temperature resolution after substantial noise suppression. Furthermore, while suppressing noise, this application fully preserves the dynamic characteristics of the target signal without affecting the system response speed, making it suitable for real-time, high-precision temperature monitoring scenarios.

Description

Technical Field

[0001] This application relates to the field of semiconductor detection technology, and in particular to a temperature measurement method and module thereof, an infrared thermometer, equipment and storage medium. Background Technology

[0002] In semiconductor manufacturing, infrared thermometers are integrated into the process chambers of equipment such as etching, thin film deposition, and rapid thermal processing to perform non-contact temperature monitoring of wafers. The accuracy of their temperature control directly determines the device performance and yield.

[0003] To address the degradation in temperature measurement performance caused by the harsh environment within the chamber, including high temperatures, plasma, and the deposition of reaction byproducts, traditional techniques employ a bypass-type fixed reference source scheme. This involves placing an independent, wavelength-stable infrared reference light source next to the main detector's optical path. The reference light is guided into the detector via a separate optical fiber or a fixed reflector, forming a reference optical path that is physically separated from the main optical path used to measure the radiation from the target. During calibration, the light source is illuminated, and the system calibrates the detector's response based on its known intensity.

[0004] In the above scheme, because a non-common path design is used that physically separates the reference light from the main optical path, the reference light does not pass through the process cavity viewport. Therefore, it is completely unaware of the transmittance reduction and additional optical noise caused by window deposition and fogging. In semiconductor processes, the most significant factor contributing to performance degradation is the contamination and degradation of the viewport. Consequently, this calibration is not very effective in improving the system's effective temperature resolution (NETD) accuracy under real-world operating conditions. Summary of the Invention

[0005] The purpose of this application is to provide a temperature measurement method and module, an infrared thermometer, device and storage medium, to overcome the defect of low effective temperature resolution accuracy under real working conditions caused by the reference light not passing through the observation window in the traditional technology.

[0006] In a first aspect, this application proposes a temperature measurement method applicable to an in-situ monitoring system for semiconductor epitaxial growth equipment. The in-situ monitoring system includes an infrared thermometer, which comprises a reference source module and a target source module. The reference source module periodically outputs a square wave modulation signal, causing the reference infrared light source to alternately be in an on-state and off-state phase during each modulation cycle. The method includes:

[0007] Obtain the reference signals for the opening phase and the closing phase, and calculate the difference between them to obtain the first differential signal;

[0008] The target signals of the opening phase and the closing phase are acquired, and the difference between the two is calculated to obtain a second differential signal; wherein, the reference signal is generated by the reference source module based on the light signal reflected from the wafer surface, the target signal is generated by the target source module based on the light signal reflected from the wafer surface, and the target signal includes the signal component obtained by splitting the reference signal;

[0009] The fluctuation information is extracted from the first differential signal and the second differential signal respectively to obtain the first fluctuation signal and the second fluctuation signal;

[0010] The first fluctuation signal and the second fluctuation signal are used as inputs to the noise prediction model to obtain the predicted noise; based on the predicted noise, the weight vector of the noise prediction model is adaptively updated so that the predicted noise gradually approaches the target noise.

[0011] The difference between the target signal and the target noise during the shutdown phase is calculated, and the target temperature is calculated based on the difference; wherein the target temperature and the target signal are in a corresponding relationship.

[0012] In one embodiment, the method for calculating the first differential signal and the second differential signal includes:

[0013] For a modulation period, multiple sets of synchronous sampling are performed in the stable region of the on-phase to obtain multiple sets of first reference data and multiple sets of first target data;

[0014] Multiple sets of synchronous sampling are performed within the stable region of the closing phase to obtain multiple sets of second reference data and multiple sets of second target data;

[0015] Calculate the difference between the first reference data and the second reference data at the same time point to obtain multiple sets of first difference data;

[0016] Calculate the difference between the first target data and the second target data at the same time point to obtain multiple sets of second difference data;

[0017] Wherein, each of the first reference data and each of the second reference data constitutes the reference signal; each of the first target data and each of the second target data constitutes the target signal; each of the first differential data constitutes the first differential signal; and each of the second differential data constitutes the second differential signal.

[0018] In one embodiment, the step of extracting the fluctuation information from the first differential signal and the second differential signal respectively to obtain the first fluctuation signal and the second fluctuation signal includes:

[0019] The slowly varying background components in the first differential signal and the second differential signal are removed to retain the short-time fluctuation components, resulting in the first fluctuation signal and the second fluctuation signal; wherein, the slowly varying background components include DC components and low-frequency slowly varying components.

[0020] In one embodiment, removing the slowly varying background component from the first differential signal and the second differential signal to retain the short-time fluctuation component, thereby obtaining the first fluctuation signal and the second fluctuation signal, includes:

[0021] Based on a preset sliding time window, the first average value of the first differential signal is calculated, and the difference between the first differential signal and the first average value is the first fluctuation signal.

[0022] Based on a preset sliding time window, the second average value of the second differential signal is calculated, and the difference between the second differential signal and the second average value is the second fluctuation signal.

[0023] In one embodiment, the noise prediction model is a linear prediction model, and its expression is:

[0024]

[0025]

[0026]

[0027] in, The input vector; This is the first wave signal; This is the second wave signal;

[0028] This is the weight vector; , which are weighting coefficients used to weight the second fluctuation signal in order to predict the noise component unique to the target measurement channel; , which is a weighting coefficient used to weight the first fluctuation signal in order to predict the noise component common to the reference channel and the target channel; is the number of samples; T represents the transpose operation.

[0029] In one embodiment, the step of adaptively updating the weight vector of the noise prediction model based on the predicted noise, so that the predicted noise gradually approximates the target noise, includes:

[0030] The prediction error is determined based on the difference between the noise reference signal and the current predicted noise; wherein, the noise reference signal is the target signal of the shutdown phase from which fluctuation information has been extracted;

[0031] The prediction error and the first and second fluctuation signals at the corresponding time are input into the recursive least squares algorithm to calculate the gain vector used to correct the weight vector, and the weight vector and its covariance matrix are updated according to the gain vector.

[0032] The updated weight vector is used for the next prediction to gradually reduce the prediction error, so that the prediction noise gradually approaches the target noise.

[0033] Secondly, this application proposes a temperature measurement module, disposed within an infrared thermometer. The infrared thermometer is located within the in-situ monitoring system of the semiconductor epitaxial growth equipment. The infrared thermometer includes a reference source module and a target source module. The reference source module periodically outputs a square wave modulation signal, causing the reference infrared light source to alternately be in an on-state and off-state phase during each modulation cycle.

[0034] The acquisition unit is configured to acquire reference signals for the on-state and the off-state, and calculate the difference between them to obtain a first differential signal; it is also configured to acquire target signals for the on-state and the off-state, and calculate the difference between them to obtain a second differential signal; wherein the reference signal is generated by the reference source module based on the light signal reflected from the wafer surface, the target signal is generated by the target source module based on the light signal reflected from the wafer surface, and the target signal includes the signal component obtained by splitting the reference signal;

[0035] An extraction unit is used to extract the fluctuation information from the first differential signal and the second differential signal respectively, to obtain the first fluctuation signal and the second fluctuation signal;

[0036] The prediction unit is used to take the first fluctuation signal and the second fluctuation signal as input to the noise prediction model to obtain the predicted noise; and adaptively update the weight vector of the noise prediction model according to the predicted noise so that the predicted noise gradually approaches the target noise.

[0037] The calculation unit is used to calculate the difference between the target signal and the target noise during the shutdown phase, and to calculate the target temperature based on the difference; wherein the target temperature and the target signal are in a corresponding relationship.

[0038] Thirdly, this application also provides an infrared thermometer, suitable for in-situ monitoring systems of semiconductor epitaxial growth equipment, including:

[0039] Reference source module, used to periodically output a reference infrared light source to the wafer surface;

[0040] The target source module is used to output the target infrared light source to the wafer surface;

[0041] The sampling module is used to split the reference reflected light signal on the wafer surface into a first light signal and a second light signal, and generate a reference signal based on the first light signal; it is also used to receive the target reflected light signal and the second light signal on the wafer surface, and generate a target signal based on the target reflected light signal and the second light signal; the modulation frequency of the reference infrared light source is higher than the modulation frequency of the target infrared light source;

[0042] The temperature measurement module described in the second aspect is used to predict target noise in the target signal based on the reference signal and the target signal, and to calculate the target temperature based on the target noise.

[0043] Fourthly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method steps of the first aspect.

[0044] Fifthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method steps of the first aspect.

[0045] The above-mentioned temperature measurement method and its module, infrared thermometer, equipment and storage medium have at least the following advantages:

[0046] The reference source module of this application periodically outputs a square wave modulation signal, causing the reference infrared light source to alternately be in the on and off phases in each modulation cycle. The reference signal and the target signal are acquired in the on and off phases of the reference infrared light source, respectively, and differentially processed. Fluctuation information is then extracted from the first differential signal and the second differential signal. The two fluctuation signals are used together as the input of the noise prediction model. The relevant fluctuation information is used to adaptively model and predict the noise of the main path system and the fluctuation of the reference light intensity. The predicted noise is gradually made closer to the real target noise through online updating of the weight vector. Finally, the target noise is canceled out from the target signal in the off phase. By adopting the above scheme, the reference optical signal and the target optical signal of this application use the same optical channel in the epitaxial growth equipment. The noise information carried in the reference signal is highly correlated with the noise in the target signal, which enables the noise prediction model to accurately model and predict the actual noise components in the main optical path. After significantly suppressing the noise, the signal-to-noise ratio and effective temperature resolution are significantly improved. In addition, while suppressing the noise, this application also fully preserves the dynamic characteristics of the target signal that reflect rapid and minute temperature changes, without affecting the system response speed, and is suitable for real-time high-precision temperature monitoring scenarios. Attached Figure Description

[0047] Figure 1This is a structural block diagram of an infrared thermometer in one embodiment;

[0048] Figure 2 This is a flowchart illustrating a temperature measurement method in one embodiment;

[0049] Figure 3 This is a flowchart illustrating the steps for calculating the first differential signal and the second differential signal in one embodiment.

[0050] Figure 4 This is a flowchart illustrating the step of updating the weight vector in one embodiment;

[0051] Figure 5 This is a structural block diagram of the temperature measurement module in one embodiment;

[0052] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0053] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0054] Some exemplary embodiments of this application have been described for illustrative purposes. It should be understood that this application may be implemented in other ways not specifically shown in the accompanying drawings.

[0055] Please see Figure 1 In one exemplary embodiment, this application provides an infrared thermometer suitable for in-situ monitoring systems of semiconductor epitaxial growth equipment. Semiconductor epitaxial growth equipment refers to process equipment used for epitaxially growing thin films or multilayer structures on a substrate, typically including metal-organic chemical vapor deposition (MOCVD), hydride vapor phase epitaxy (HVPE), molecular beam epitaxy (MBE), and various chemical vapor deposition (CVD) methods. The in-situ monitoring system is a measurement subsystem that performs real-time measurement and data processing of the epitaxial growth state of the wafer within the reaction chamber without disrupting the process environment or interrupting growth. Further, a base for placing the wafer is provided within the reaction chamber. The base is made of high-purity graphite and is supported by a rotating shaft driven by a motor, allowing it to rotate continuously during the epitaxial growth process of the wafer.

[0056] An infrared thermometer is positioned above the reaction chamber and includes a reference source module, a target source module, a sampling module, and a temperature measurement module.

[0057] The reference source module is used to periodically output a reference infrared light source to the wafer surface.

[0058] Specifically, the reference source module includes a light source and a driving circuit. The light source is used to output a reference infrared light source. In this embodiment, a vertical-cavity surface-emitting laser (VCSEL) is preferred, which has advantages such as fast modulation speed, high spectral purity, and small size. The driving circuit provides a high-frequency square wave modulation driving current to the VCSEL, causing it to output periodic infrared light pulses, so that the reference infrared light source alternately enters an on and off phase in each adjustment cycle. For example, the high-frequency range in this embodiment is 1-10kHz.

[0059] The target source module is used to output the target infrared light source to the wafer surface.

[0060] Specifically, the target source module includes a radiation source and an optical lens. The radiation source is used to output the target infrared light source; the optical lens receives the diverging beam from the radiation source, collimates, focuses and homogenizes it through its built-in optical group, and finally illuminates the target area on the wafer surface with a light spot of preset shape and energy density.

[0061] The sampling module is used to split the reference reflected light signal on the wafer surface into a first light signal and a second light signal, and generate a reference signal based on the first light signal; it is also used to receive the target reflected light signal and the second light signal on the wafer surface, and generate a target signal based on the target reflected light signal and the second light signal; the modulation frequency of the reference infrared light source is higher than the modulation frequency of the target infrared light source.

[0062] Specifically, the sampling module includes a beam splitter, a reference detector, and a main detector.

[0063] The beam splitter, tilted and positioned on the reference optical path, is designed with a coating that maximizes the transmittance (T>90%) of the target reflected light signal and maximizes the reflectance (R>90%), ensuring near-loss-free transmission of the target signal. The reference signal (S_ref) is partially reflected and merges with the target signal. This reflected light and the target signal pass through all optical elements together before reaching the main detector. In this scheme, the target and reference signals follow identical optical paths. This identical path means they share the same contaminated observation window, the same lens group, and the same airflow disturbances. Therefore, the noise information carried by the reference signal is highly correlated with the noise in the target signal. This allows the subsequent adaptive algorithm to model and predict the true noise of the main optical path with extreme accuracy, resulting in optimal signal-to-noise ratio improvement and effective temperature resolution (NETD) optimization after noise reduction.

[0064] A reference detector is used to receive a first optical signal and generate a reference signal based on the first optical signal; wherein the first optical signal is obtained by splitting a reference reflected optical signal from the wafer surface, and the reference reflected optical signal is generated by the wafer surface reflecting a reference infrared light source.

[0065] The main detector is used to receive the second optical signal and the target reflected optical signal, and to generate a target signal based on the second optical signal and the target reflected optical signal.

[0066] In this design, the sampling clocks of the reference detector and the main detector are synchronized, and the modulation frequency of the reference infrared light source is higher than that of the target infrared light source. By employing this scheme, and using a modulation frequency significantly higher than the target signal refresh rate, the injection of the reference signal, the extraction of noise samples, and noise reduction processing can be completed within the microscopic timescale of normal temperature measurement. This allows this application to leap from a periodically interrupted offline calibration mode to a mode where real-time correction is performed synchronously with each normal temperature measurement. The effective temperature resolution is optimized in every measurement, thoroughly meeting the stringent requirements of continuous and seamless monitoring in industrial scenarios such as semiconductor manufacturing.

[0067] Furthermore, by eliminating all moving mechanical parts, the aforementioned solution fundamentally eliminates particulate contaminants generated by the opening and closing of the mechanical shutter, fully meeting the ultra-cleanliness requirements of semiconductor process chambers. Simultaneously, the absence of wear and jamming risks significantly extends the device's lifespan and mean time between failures (MTBF), while also reducing maintenance costs and machine downtime risks. Moreover, the newly added reference source module, spectrometer, and reference detector are all miniaturized, solid-state electronic devices, eliminating the need for complex mechanical structures and drive mechanisms. This results in a small-sized, low-power device that is easily integrated into existing or newly designed infrared temperature measurement systems. While improving performance, it effectively controls manufacturing costs and integration complexity, possessing extremely high practical value and market potential.

[0068] The temperature measurement module is used to predict target noise in the target signal based on a reference signal and a target signal, and to calculate the target temperature based on the target noise. Specifically, the temperature measurement module acquires reference signals for the on and off phases, calculates the difference between them to obtain a first differential signal; acquires target signals for the on and off phases, calculates the difference between them to obtain a second differential signal; wherein the reference signal is generated by the reference source module based on the reflected light signal from the wafer surface, and the target signal is generated by the target source module based on the reflected light signal from the wafer surface, and the target signal includes the signal component obtained by splitting the reference signal; the fluctuation information in the first and second differential signals is extracted respectively to obtain a first fluctuation signal and a second fluctuation signal; the first and second fluctuation signals are used as inputs to a noise prediction model to obtain predicted noise; based on the predicted noise, the weight vector of the noise prediction model is adaptively updated so that the predicted noise gradually approaches the target noise; the difference between the target signal and the target noise in the off phase is calculated, and the target temperature is calculated based on this difference; wherein the target temperature and the target signal are in a corresponding relationship.

[0069] The aforementioned infrared thermometer periodically outputs a square wave modulation signal, causing the reference infrared light source to alternately be in an on-state and off-state phase within each modulation cycle. Reference and target signals are acquired during the on-state and off-state phases of the reference infrared light source, respectively, and differentially processed. Fluctuation information is then extracted from the first and second differential signals. These two fluctuation signals are used together as input to a noise prediction model. The relevant fluctuation information is used to adaptively model and predict the noise of the main path system and the fluctuation of the reference light intensity. Online updates of the weight vector gradually bring the predicted noise closer to the actual target noise. Finally, the target noise is canceled out from the target signal during the off-state phase. By adopting the above scheme, the reference optical signal and the target optical signal of this application use the same optical channel in the epitaxial growth equipment. The noise information carried in the reference signal is highly correlated with the noise in the target signal, which enables the noise prediction model to accurately model and predict the actual noise components in the main optical path. After significantly suppressing the noise, the signal-to-noise ratio and effective temperature resolution are significantly improved. In addition, while suppressing the noise, this application also fully preserves the dynamic characteristics of the target signal that reflect rapid and minute temperature changes, without affecting the system response speed, and is suitable for real-time high-precision temperature monitoring scenarios.

[0070] Please see Figure 2 In one exemplary embodiment, this application provides a temperature measurement method applicable to an in-situ monitoring system for semiconductor epitaxial growth equipment. The in-situ monitoring system includes an infrared thermometer and specifically includes the following steps:

[0071] Step 202: Obtain the reference signals for the on and off phases, and calculate the difference between them to obtain the first differential signal.

[0072] Specifically, during the measurement process, the reference detector continuously acquires reference reflected light signals from the wafer surface at a preset sampling frequency. Within the stable time window of the on and off phases of each modulation cycle, multiple sets of sampling points are selected as reference signals. For example, let the reference signal in the on phase be M_on, and the reference signal in the off phase be M_off. Then, the first differential signal ΔM = M_on - M_off, where the value of M_on is proportional to the second optical signal, and the theoretical value of M_off is zero.

[0073] Step 204: Obtain the target signals for the on and off phases, and calculate the difference between them to obtain the second differential signal; wherein, the reference signal is generated by the reference source module based on the reflected light signal from the wafer surface, the target signal is generated by the target source module based on the reflected light signal from the wafer surface, and the target signal includes the signal component obtained by splitting the reference signal.

[0074] Specifically, during the measurement process, the main detector continuously acquires the target reflected light signal and the second light signal from the wafer surface at a preset sampling frequency. Within the stable time window of the on and off phases of each modulation cycle, multiple sets of sampling points are selected as the target signal. For example, the target signal in the on phase is denoted as S_on = S_target + S_ref + N_system, where S_target is the target signal corresponding to the target reflected light signal, S_ref is the target signal corresponding to the second light signal, and N_system is the system noise. The target signal in the off phase is denoted as S_off = S_target + N_system, then the second differential signal ΔS = S_on - S_off = S_ref + ΔN_system. It should be noted that the purpose of the above differential is to extract the pure reference signals ΔS and ΔM from the mixed signal for subsequent analysis. Since the system noise is a random process, the noise in the on and off phases is independent; therefore, the differential ΔS still contains the influence of noise.

[0075] Step 206: Extract the fluctuation information from the first differential signal and the second differential signal respectively to obtain the first fluctuation signal and the second fluctuation signal.

[0076] Specifically, the fluctuation information in the first differential signal refers to the fluctuation in the reference path, characterizing the intensity fluctuation of the reference infrared source itself. The fluctuation information in the second differential signal refers to the fluctuation in the main path, characterizing the intensity fluctuation of the reference infrared source and the fluctuation of system noise in the main path.

[0077] Furthermore, in this embodiment, the fluctuation information of the target signal S_off during the shutdown phase is also extracted. This fluctuation information represents the short-term fluctuation of S_off. Since S_target remains basically unchanged within milliseconds, the fluctuation information of S_off can be regarded as a real-time sample of the noise in the main path system.

[0078] Step 208: Use the first and second wave signals as inputs to the noise prediction model to obtain the predicted noise; based on the predicted noise, adaptively update the weight vector of the noise prediction model so that the predicted noise gradually approaches the target noise.

[0079] Specifically, the noise prediction model is an online prediction model. At each sampling moment, the current first and second fluctuation signals are used as inputs, and the predicted noise at the current sampling moment is output. Then, based on the prediction error between the noise sample and the predicted noise at the current sampling moment, the weight vector of the noise prediction model is adaptively updated, so that the model parameters converge over time. Thus, the main path noise is predicted and tracked online on a sample-by-sample basis throughout the measurement process.

[0080] Step 210: Calculate the difference between the target signal and the target noise during the shutdown phase, and calculate the target temperature based on the difference; wherein the target temperature and the target signal are in a corresponding relationship.

[0081] Specifically, the target signal S_off during the shutdown phase is S_target + N_system. The predicted target noise is subtracted from the original signal to obtain the denoised real signal. The final temperature value can then be calculated based on this real signal.

[0082] Generally, under conditions of a fixed wavelength, known emissivity, optical channel, and detector response, the wafer surface temperature exhibits a monotonic relationship with the radiation intensity of that wavelength; that is, as the temperature increases, the radiation signal in the corresponding wavelength increases. In practical engineering applications, this relationship can be characterized using spectral signal-temperature curves or tables.

[0083] The aforementioned temperature measurement method periodically outputs a square wave modulation signal, causing the reference infrared light source to alternately be in an on-state and off-state phase within each modulation cycle. Reference and target signals are acquired during the on-state and off-state phases of the reference infrared light source, respectively, and differentially processed. Fluctuation information is then extracted from the first and second differential signals. Both fluctuation signals are used as input to a noise prediction model. The relevant fluctuation information is used to adaptively model and predict the noise of the main path system and the fluctuation of the reference light intensity. Online updates of the weight vector gradually bring the predicted noise closer to the actual target noise. Finally, the target noise is canceled out from the target signal during the off-state phase. By adopting the above scheme, the reference optical signal and the target optical signal of this application use the same optical channel in the epitaxial growth equipment. The noise information carried in the reference signal is highly correlated with the noise in the target signal, which enables the noise prediction model to accurately model and predict the actual noise components in the main optical path. After significantly suppressing the noise, the signal-to-noise ratio and effective temperature resolution are significantly improved. In addition, while suppressing the noise, this application also fully preserves the dynamic characteristics of the target signal that reflect rapid and minute temperature changes, without affecting the system response speed, and is suitable for real-time high-precision temperature monitoring scenarios.

[0084] Please see Figure 3 Optionally, the methods for calculating the first differential signal and the second differential signal include:

[0085] Step 302: For a modulation period, perform multiple sets of synchronous sampling in the stable region of the start phase to obtain multiple sets of first reference data and multiple sets of first target data.

[0086] Step 304: Perform multiple sets of synchronous sampling within the stable region of the shutdown phase to obtain multiple sets of second reference data and multiple sets of second target data.

[0087] Step 306: Calculate the difference between the first reference data and the second reference data corresponding to the same time point to obtain multiple sets of first difference data.

[0088] Step 308: Calculate the difference between the first target data and the second target data at the same time point to obtain multiple sets of second differential data. Each first reference data and each second reference data constitutes a reference signal; each first target data and each second target data constitutes a target signal; each first differential data constitutes a first differential signal; and each second differential data constitutes a second differential signal.

[0089] Specifically, in this embodiment of the application, within one adjustment cycle, the reference signal and the target signal are synchronously sampled multiple times during the on-phase and off-phase. That is, at each sampling moment during the on-phase, the first optical signal in the reference channel, the second optical signal in the main channel, and the target reflected optical signal are simultaneously acquired to obtain a set of first reference data and first target data. Similarly, at each sampling moment during the off-phase, the first optical signal in the reference channel, the second optical signal in the main channel, and the target reflected optical signal are simultaneously acquired to obtain a set of second reference data and second target data. It should be understood that the first and second optical signals are theoretically zero during the off-phase.

[0090] By employing the above scheme, which involves multiple synchronous samplings of the reference and target signals within a modulation period, compared to sampling only at a single moment, the approach can significantly suppress instantaneous random noise and single sampling errors by averaging, filtering, or removing outliers from multiple sets of differential data. This reduces the variance of the differential signal and improves the stability and reliability of the differential results between the reference and target signals. Furthermore, since each set of data is acquired synchronously at the same moment, the correspondence between the reference and target channels under the same optical state can be more accurately characterized. This provides richer and more consistent sample data for subsequent adaptive noise modeling based on the first and second wave signals, thereby improving the convergence speed and modeling accuracy of the noise prediction model and further enhancing the overall signal-to-noise ratio and temperature resolution of the temperature measurement.

[0091] Optionally, fluctuation information is extracted from the first differential signal and the second differential signal respectively to obtain the first fluctuation signal and the second fluctuation signal, including:

[0092] The slowly varying background components in the first differential signal and the second differential signal are removed to retain the short-time fluctuation components, resulting in the first fluctuation signal and the second fluctuation signal; wherein, the slowly varying background components include DC components and low-frequency slowly varying components.

[0093] Specifically, the slowly varying background component refers to the signal components that change slowly over multiple modulation cycles and are used to characterize the signal baseline and its low-frequency drift. These include DC components and low-frequency slowly varying components introduced by factors such as system bias, optical channel temperature drift, and window contamination.

[0094] Short-time fluctuation components refer to signal components that change rapidly over multiple sampling periods or multiple modulation periods, exhibiting zero-mean fluctuations relative to local averages. These components are primarily caused by rapid fluctuations in the intensity of the reference light source, random noise in the detector and signal link, and rapid temperature changes in the object under test, exhibiting high frequency characteristics. In this embodiment, the short-time fluctuation component characterizes the instantaneous fluctuation features in both the target signal and the reference signal.

[0095] Optionally, the slowly varying background component is removed from the first differential signal and the second differential signal to retain the short-time fluctuation component, resulting in the first fluctuation signal and the second fluctuation signal, including:

[0096] Based on a preset sliding time window, the first average value of the first differential signal is calculated, and the difference between the first differential signal and the first average value is the first fluctuation signal.

[0097] Based on a preset sliding time window, the second average value of the second differential signal is calculated, and the difference between the second differential signal and the second average value is the second fluctuation signal.

[0098] Specifically, during continuous sampling, the first differential signal and the second differential signal obtained in each modulation period are stored in the buffer in chronological order, and the slowly varying background components in the differential signal are estimated and removed by using a sliding time window.

[0099] For the first differential signal, at each sampling time, a time window of preset length is formed by selecting multiple sampling points before and after the current sampling point. The first differential signal within the window is averaged and summed, and the resulting first average value is used as the slowly varying background component corresponding to that sampling point. Then, the slowly varying background component is subtracted from the first differential signal at the current sampling point to obtain the corresponding first short-time fluctuation component. The short-time fluctuation components of all sampling points constitute the first fluctuation signal. It should be understood that, in addition to calculating the average sum, other processing can be performed as needed. For example, a low-order polynomial fitting can be performed on the first differential signal within the window, and the fitting result can be used as the slowly varying background component corresponding to that sampling point.

[0100] For the second differential signal, the same sliding time window and smoothing method as the first differential signal are used to estimate its slowly varying background component. The second differential signal at each sampling point is subtracted from the corresponding slowly varying background component to obtain the second short-time fluctuation component. The short-time fluctuation components of all sampling points constitute the second fluctuation signal.

[0101] By employing the above scheme, and performing sliding window smoothing and background subtraction on the first and second differential signals in the time domain, the slowly changing background components caused by optical window contamination, optical element temperature drift, and baseline drift can be separated from the differential signals. Only short-term fluctuation components closely related to reference light intensity fluctuations and system noise are retained, thus allowing the first and second fluctuation signals to focus more on characterizing instantaneous noise characteristics. This reduces the interference of slow drift on the adaptive noise prediction model, preventing the model parameters from being pulled by long-term trends and making it difficult to converge in a timely manner. Furthermore, it improves the signal-to-noise ratio of the noise prediction input, which helps the model to more accurately learn the correlation between the main path noise and the reference path fluctuation, thereby improving the accuracy and stability of noise prediction and further enhancing the real-time signal-to-noise ratio and effective temperature resolution in the temperature measurement process.

[0102] Optionally, the noise prediction model is a linear prediction model, and its expression is:

[0103]

[0104]

[0105]

[0106] in, The input vector; This is the first wave signal; This is the second wave signal; This is the weight vector; These are weighting coefficients used to weight the second fluctuation signal in order to predict the noise component unique to the target measurement channel; These are weighting coefficients used to weight the first fluctuation signal in order to predict the noise component common to the reference channel and the target channel. is the number of samples; T represents the transpose operation.

[0107] Please see Figure 4 Optionally, the weight vector of the noise prediction model is adaptively updated based on the predicted noise, so that the predicted noise gradually approximates the target noise, including:

[0108] Step 402: Determine the prediction error based on the difference between the noise reference signal and the current predicted noise; wherein, the noise reference signal is the target signal of the off-phase from which fluctuation information has been extracted.

[0109] Step 404: Input the prediction error and the first and second fluctuation signals at the corresponding time into the recursive least squares algorithm to calculate the gain vector used to correct the weight vector, and update the weight vector and its covariance matrix according to the gain vector.

[0110] Step 406: Use the updated weight vector for the next prediction to gradually reduce the prediction error and make the prediction noise gradually approach the target noise.

[0111] Specifically, the purpose of this application is to construct a linear prediction model that utilizes noise-related auxiliary inputs. and The noise of the main path is predicted. The predicted noise is subtracted from the original signal in real time to obtain the denoised real target signal. This real target signal is used to calculate the final temperature value, and its expression is: S_final[k] = S_off[k]- N_predicted[k]; where S_off[k] is the target signal in the off stage and N_predicted[k] is the target noise.

[0112] Furthermore, in order to make the predicted noise infinitely close to the real noise, the coefficients of the noise prediction model need to be continuously updated. The steps to achieve this include:

[0113] In each sampling period, the prediction error is calculated, and its expression is: e[k] = N_actual[k] - N_predicted[k]; where N_actual[k] is the obtained real noise reference value. Since the fluctuation information of S_off is directly derived from the main path and contains almost no change in the target signal, in this embodiment, the fluctuation information of S_off δS_off[k] is used to approximate N_actual[k].

[0114] The Recursive Least Squares (RLS) algorithm is used to update the weights of the noisy prediction model. Specifically, the RLS algorithm uses a set of recursive formulas to efficiently update the weights using the current prediction error and the input signal, minimizing the long-term average sum of squared prediction errors. Its core formula is as follows:

[0115] Formula (1);

[0116] Formula (2);

[0117] Formula (3);

[0118] Formula (1) above is the gain vector expression, formula (2) is the weight update formula, and formula (3) is the covariance update expression. Wherein, The Kalman gain vector determines the magnitude of the coefficient correction caused by the current error. The inverse correlation matrix of the input signal is estimated and is a state variable within the algorithm. The forgetting factor has a value range of 0 < λ ≤ 1. In this embodiment, it is set to 0.99 to give new data a higher weight, enabling the algorithm to track non-stationary noise with slowly changing statistical characteristics.

[0119] The above steps are repeated for each sampling period k, thereby achieving uninterrupted online real-time noise reduction, and ultimately making the predicted noise gradually approach the target noise.

[0120] It should be noted that the recursive least squares method is used to update the model coefficients in this embodiment. In practical applications, an appropriate adaptive algorithm with online learning capabilities can be selected as needed.

[0121] Unlike traditional noise reduction methods that rely on software smoothing such as low-pass filtering, the adaptive noise cancellation technology of this application actively processes the signal, precisely subtracting the predicted noise component from the main signal, rather than simply smoothing the entire signal. Therefore, while effectively suppressing noise, this application perfectly preserves the rapid, subtle real temperature changes inherent in the target signal, ensuring that the system does not sacrifice response speed or distortion when monitoring rapid thermal processing processes such as Rapid Thermal Annealing (RTP). Furthermore, an adaptive algorithm based on recursive least squares continuously and recursively updates the model coefficients. This makes the noise reduction system not a filter with fixed parameters, but an intelligent agent with "learning" capabilities. It can automatically track and adapt to changes in noise characteristics caused by system aging, lens contamination accumulation, and slow temperature drift, ensuring optimal noise reduction and temperature resolution throughout the entire equipment lifecycle.

[0122] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0123] Based on the same inventive concept, this application also provides a temperature measurement module. This system is applicable to the above-described temperature measurement method. The solution provided by this system is similar to the solution described in the above-described method. Therefore, the specific limitations in one or more embodiments provided below can be found in the limitations of the method described above, and will not be repeated here.

[0124] Please see Figure 5 In one embodiment, this application provides a temperature measurement module, which is disposed in the infrared thermometer provided in the above embodiment, and includes: an acquisition unit, an extraction unit, a prediction unit, and a calculation unit.

[0125] The acquisition unit is used to acquire reference signals during the on and off phases and calculate the difference between them to obtain a first differential signal; it is also used to acquire target signals during the on and off phases and calculate the difference between them to obtain a second differential signal; wherein the reference signal is generated by the reference source module based on the light signal reflected from the wafer surface, the target signal is generated by the target source module based on the light signal reflected from the wafer surface, and the target signal includes the signal component obtained by splitting the reference signal.

[0126] The extraction unit is used to extract the fluctuation information from the first differential signal and the second differential signal respectively, so as to obtain the first fluctuation signal and the second fluctuation signal.

[0127] The prediction unit is used to take the first and second wave signals as inputs to the noise prediction model to obtain the predicted noise; based on the predicted noise, it adaptively updates the weight vector of the noise prediction model so that the predicted noise gradually approaches the target noise.

[0128] The calculation unit is used to calculate the difference between the target signal and the target noise during the shutdown phase, and to calculate the target temperature based on the difference; wherein the target temperature and the target signal are in a corresponding relationship.

[0129] Optionally, the acquisition unit calculates the first differential signal and the second differential signal in the following manner: for one modulation period, performing multiple sets of synchronous sampling in the stable region of the on-phase to obtain multiple sets of first reference data and multiple sets of first target data; performing multiple sets of synchronous sampling in the stable region of the off-phase to obtain multiple sets of second reference data and multiple sets of second target data; calculating the difference between the first reference data and the second reference data at the same time point to obtain multiple sets of first differential data; calculating the difference between the first target data and the second target data at the same time point to obtain multiple sets of second differential data. Wherein, each first reference data and each second reference data constitutes a reference signal; each first target data and each second target data constitutes a target signal; each first differential data constitutes a first differential signal; and each second differential data constitutes a second differential signal.

[0130] Optionally, the extraction unit extracts the fluctuation information from the first differential signal and the second differential signal respectively to obtain the first fluctuation signal and the second fluctuation signal, including: removing the slowly varying background component from the first differential signal and the second differential signal to retain the short-time fluctuation component, thereby obtaining the first fluctuation signal and the second fluctuation signal; wherein, the slowly varying background component includes a DC component and a low-frequency slowly varying component.

[0131] Optionally, the extraction unit removes the slowly varying background components from the first differential signal and the second differential signal to retain the short-time fluctuation components, thereby obtaining the first fluctuation signal and the second fluctuation signal. This includes: calculating a first average value of the first differential signal based on a preset sliding time window, wherein the difference between the first differential signal and the first average value is the first fluctuation signal; and calculating a second average value of the second differential signal based on a preset sliding time window, wherein the difference between the second differential signal and the second average value is the second fluctuation signal.

[0132] Optionally, the noise prediction model is a linear prediction model, and its expression is:

[0133]

[0134]

[0135]

[0136] in, The input vector; This is the first wave signal; This is the second wave signal; This is the weight vector; These are weighting coefficients used to weight the second fluctuation signal in order to predict the noise component unique to the target measurement channel; These are weighting coefficients used to weight the first fluctuation signal in order to predict the noise component common to the reference channel and the target channel. is the number of samples; T represents the transpose operation.

[0137] Optionally, the prediction unit adaptively updates the weight vector of the noise prediction model based on the predicted noise, so that the predicted noise gradually approaches the target noise. This includes: determining the prediction error based on the difference between the noise reference signal and the current predicted noise; wherein the noise reference signal is the target signal of the off-phase with extracted fluctuation information; inputting the prediction error and the first and second fluctuation signals at the corresponding time into a recursive least squares algorithm to calculate the gain vector used to correct the weight vector; updating the weight vector and its covariance matrix based on the gain vector; and using the updated weight vector for the next prediction to gradually reduce the prediction error and make the predicted noise gradually approach the target noise.

[0138] The aforementioned temperature measurement module acquires reference and target signals during the on and off phases of the reference infrared light source, respectively, and performs differential processing on them. Fluctuation information is then extracted from the first and second differential signals. Both fluctuation signals are used as input to a noise prediction model. The relevant fluctuation information is used to adaptively model and predict the noise of the main path system and the fluctuation of the reference light intensity. Online updates of the weight vector gradually bring the predicted noise closer to the actual target noise. Finally, the target noise is canceled out from the target signal during the off phase. Using this scheme, the reference and target light signals in this application utilize the same optical channel in the epitaxial growth equipment. The noise information carried in the reference signal is highly correlated with the noise in the target signal, enabling the noise prediction model to accurately model and predict the actual noise components in the main optical path. This significantly improves the signal-to-noise ratio and effective temperature resolution after suppressing noise. Furthermore, while suppressing noise, this application also fully preserves the dynamic characteristics of the target signal reflecting rapid and minute temperature changes, without affecting the system response speed, making it suitable for real-time, high-precision temperature monitoring scenarios.

[0139] Each module in the aforementioned temperature measurement module can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0140] In one feasible embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the aforementioned temperature measurement method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0141] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment to which the present application is applied. Specific computer equipment may include, for example, [the following is a list of possible additional structures]. Figure 6 The diagram shows more or fewer components, or combinations of certain components, or different component arrangements.

[0142] In one feasible embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method steps in the above-described temperature measurement method.

[0143] In one feasible embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method steps of the temperature measurement method described above.

[0144] In one feasible embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the method steps in the temperature measurement method described above.

[0145] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0146] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0147] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A temperature measurement method, characterized by, The application relates to an in-situ monitoring system for a semiconductor epitaxial growth device, the in-situ monitoring system comprising an infrared temperature detector, the infrared temperature detector comprising a reference source module and a target source module, the reference source module periodically outputting a square wave modulation signal, so that a reference infrared light source alternately enters an opening stage and a closing stage in each modulation period; the method comprises the following steps: acquiring reference signals in the opening stage and the closing stage, calculating the difference between the two to obtain a first differential signal; acquiring target signals in the opening stage and the closing stage, calculating the difference between the two to obtain a second differential signal; wherein the reference signal is generated by the reference source module based on a wafer surface reflected light signal, the target signal is generated by the target source module based on a wafer surface reflected light signal, and the target signal comprises a signal component obtained by splitting the reference signal; extracting fluctuation information in the first differential signal and the second differential signal respectively to obtain a first fluctuation signal and a second fluctuation signal; taking the first fluctuation signal and the second fluctuation signal as inputs of a noise prediction model to obtain a predicted noise; according to the predicted noise, the weight vector of the noise prediction model is adaptively updated, so that the predicted noise gradually approximates to a target noise; calculating the difference between the target signal in the closing stage and the target noise, and calculating a target temperature according to the difference; wherein the target temperature and the target signal have a corresponding relationship.

2. The method of claim 1, wherein, The way of calculating the first differential signal and the second differential signal comprises the following steps: for one modulation period, a plurality of groups of synchronous sampling are performed in the stable region of the opening stage to obtain a plurality of groups of first reference data and a plurality of groups of first target data; a plurality of groups of synchronous sampling are performed in the stable region of the closing stage to obtain a plurality of groups of second reference data and a plurality of groups of second target data; the difference between the first reference data and the second reference data corresponding to the same time point is calculated respectively to obtain a plurality of groups of first differential data; the difference between the first target data and the second target data corresponding to the same time point is calculated respectively to obtain a plurality of groups of second differential data; wherein each of the first reference data and each of the second reference data constitute the reference signal; each of the first target data and each of the second target data constitute the target signal; each of the first differential data constitutes the first differential signal; and each of the second differential data constitutes the second differential signal.

3. The method of claim 1, wherein, The method for extracting fluctuation information in the first differential signal and the second differential signal respectively to obtain a first fluctuation signal and a second fluctuation signal comprises the following steps: the slowly-varying background components in the first differential signal and the second differential signal are removed to retain short-time fluctuation components, so as to obtain the first fluctuation signal and the second fluctuation signal; wherein the slowly-varying background components comprise a direct current component and a low-frequency slow-varying component.

4. The method of claim 3, wherein, The method for removing the slowly-varying background components in the first differential signal and the second differential signal to retain short-time fluctuation components, so as to obtain the first fluctuation signal and the second fluctuation signal, comprises the following steps: a first average value of the first differential signal is calculated based on a preset sliding time window, and the difference between the first differential signal and the first average value is the first fluctuation signal; The second average value of the second difference signal is calculated based on a preset sliding time window, and a difference value of the second difference signal relative to the second average value is the second fluctuation signal.

5. The method of claim 1, wherein, The noise prediction model is a linear prediction model, and an expression of the noise prediction model is: wherein is an input vector; is a first fluctuation signal; is a second fluctuation signal; is a weight vector; is a weight coefficient for weighting the second fluctuation signal to predict a noise component unique to the target measurement channel; is a weight coefficient for weighting the first fluctuation signal to predict a noise component common to the reference channel and the target channel; is the number of samples; T denotes the transpose operation.

6. The method of claim 5, wherein, The weight vector of the noise prediction model is adaptively updated according to the predicted noise, so that the predicted noise gradually approaches the target noise, and the method comprises the following steps: A prediction error is determined according to a difference between a noise reference signal and a current predicted noise, wherein the noise reference signal is a target signal of an off stage in which fluctuation information is extracted; The prediction error, the first fluctuation signal and the second fluctuation signal at the corresponding moment are input into a recursive least square algorithm, and a gain vector for correcting the weight vector is calculated, and the weight vector and a covariance matrix thereof are updated according to the gain vector; The updated weight vector is used for next prediction, so as to gradually reduce the prediction error, so that the predicted noise gradually approaches the target noise.

7. A temperature measurement module, characterized by The infrared thermometer is arranged in the in-situ monitoring system of the semiconductor epitaxial growth equipment, and the infrared thermometer comprises a reference source module and a target source module. The reference source module periodically outputs a square wave modulation signal, so that the reference infrared light source alternately enters an on stage and an off stage in each modulation period. The infrared thermometer comprises: An acquisition unit is configured to acquire reference signals of the on stage and the off stage, calculate a difference between the reference signals to obtain a first difference signal, and acquire target signals of the on stage and the off stage, calculate a difference between the target signals to obtain a second difference signal. The reference signals are generated by the reference source module based on wafer surface reflected light signals, and the target signals are generated by the target source module based on wafer surface reflected light signals. The target signals comprise a signal component obtained by splitting the reference signals. A extraction unit is configured to extract fluctuation information in the first difference signal and the second difference signal respectively to obtain a first fluctuation signal and a second fluctuation signal. A prediction unit is configured to take the first fluctuation signal and the second fluctuation signal as inputs of a noise prediction model to obtain a predicted noise, and adaptively update a weight vector of the noise prediction model according to the predicted noise, so that the predicted noise gradually approaches a target noise. A calculation unit is configured to calculate a difference between the target signals of the off stage and the target noise, and calculate a target temperature according to the difference. The target temperature and the target signals have a corresponding relationship.

8. An infrared thermometer, characterized by The in-situ monitoring system is suitable for a semiconductor epitaxial growth equipment, and comprises: A reference source module is configured to periodically output a reference infrared light source to a wafer surface. A target source module is configured to output a target infrared light source to a wafer surface. A sampling module is configured to split the reference reflected light signal of the wafer surface into a first light signal and a second light signal, and generate a reference signal according to the first light signal; and receive a target reflected light signal of the wafer surface and the second light signal, and generate a target signal according to the target reflected light signal and the second light signal; the modulation frequency of the reference infrared light source is higher than the modulation frequency of the target infrared light source; The temperature measurement module as claimed in claim 7 is configured to predict a target noise in the target signal according to the reference signal and the target signal, and calculate a target temperature according to the target noise. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method of any one of claims 1-6.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1-6.

Images