PPG multi-channel spectrum diffuse reflection food baking degree detection method

Abstract

The invention provides a PPG multi-channel spectrum diffuse reflection food baking degree detection method, which comprises the following steps: S1, selecting a plurality of monochromatic light sources or full-wave band light sources and detectors, and arranging dodging sheets on the light sources and the detectors in a matching manner; the light source is utilized to emit light of a required wave band, and the light is scattered by the light source dodging sheet, so that the light is uniformly irradiated on a sample to be detected; s2, the light rays are emitted after passing through the sample, and judgment is carried out according to the absorptivity and the reflectivity of the light rays with different wavelengths irradiated on the sample; the reflectivity is the ratio of the intensity of light which can be reflected back when the light irradiates the sample to the irradiation intensity; s3, the reflected light is dodging again through a dodging sheet of the detector and then enters the detector; the detector detects the reflectivity of light rays with different wavelengths; real-time detection and quantitative analysis of the baking degree of the food are realized by collecting multi-channel spectral data reflected by the surface of the food and combining proper data processing and algorithms.

Description

Technical Field

[0001] The present invention relates to the technical field of food roasting degree detection methods, and in particular to a PPG multi-channel spectral diffuse reflectance food roasting degree detection method. Background Art

[0002] Roasting degree is a parameter created by Agtron to detect the roasting degree of food, such as cereals, bread, beans and other foods. In specific applications, there are currently two main problems in this field: 1. Continuous spectrum detection is costly: The equipment cost of using the continuous spectrum method for roasting degree detection is relatively high because this method requires expensive sensor costs. The current desktop machine uses a multi-spectral continuous spectrum method, which is measured by white light grating spectroscopy and an integrating sphere.

[0003] 2. Portable roasting degree testers have large errors: Existing portable roasting degree testers use a single infrared light source (850nm wavelength) for detection, which may cause large errors when detecting light colors, and the presence of a single light source during the measurement process is easily affected by external factors.

[0004] Based on these issues, how to effectively solve the problems of high cost and large error in traditional methods, so as to achieve more accurate and economical food baking degree detection, is a problem developed and considered by technical personnel in this field. Summary of the invention

[0005] The problems existing in the prior art solved by this application are: Traditional food roasting degree detection uses continuous spectrum detection and a single infrared light source for detection, which has the problems of high cost and large error.

[0006] The solution to the technical problem of the present invention is: A PPG multi-channel spectral diffuse reflectance food roasting degree detection method is provided, comprising the following steps: S1: Select multiple monochromatic light sources or full-band light sources and detectors, and each light source and detector is matched with a homogenizer; use the light source to emit light of the required band, and disperse the light through the light source homogenizer, so that the light is evenly irradiated on the sample to be tested; S2: The light is emitted after passing through the sample, and the judgment is made based on the absorptivity and reflectivity of light of different wavelengths irradiated on the sample; the reflectivity is the ratio of the intensity of light reflected back from the sample to the intensity of irradiation; S3: The reflected light is homogenized again by the detector homogenizer, and then enters the detector; the detector detects the reflectivity of light of different wavelengths, and the reflection reflected on the detector is the reflection intensity; the higher the reflectivity of the wavelength, the stronger the signal can be obtained in the detector, and vice versa, the signal of the wavelength with a smaller reflectivity detected by the detector is weaker; S4: The detector converts the detected signal into a digital signal through an IV amplifier or a digital analog acquisition front end, puts it into a register, and obtains the final agtron value through the algorithm of the main control circuit board.

[0007] Preferably, in step S4, a multivariate linear regression model method is used to calculate the agtron value; the spectral intensities of five bands are preset, expressed as 𝑥1, 𝑥2, 𝑥3, 𝑥4, and 𝑥5, wherein each 𝑥𝑖 is a predictor, representing the spectral intensity of a specific band; the agtron value is 𝑦; the multivariate linear regression model is expressed as y=β0+β1x1+β2x2+β3x3+β4x4+β5x5+ϵ; wherein 𝛽0 is the intercept term; β1, β2, β3, β4, and β5 are model parameters, representing the weight of the influence of each band on the agtron value; ϵ is an error term, used to capture variations that cannot be explained by the model.

[0008] Preferably, in step S4, the least square method is used to estimate the model parameter β value; the goal is to minimize the residual sum of squares, that is, the sum of squares of the difference between the actual agtron value and the agtron value predicted by the model; the parameter estimation formula is: β=(X T X) -1 X T y; X is the design matrix, containing all spectral intensity data and a constant term column (corresponding to the intercept term); Y is a vector containing all observed agtron values; β^ is a vector of parameter estimates; After obtaining the model parameters, the new spectral data is input into the model to calculate the predicted agtron value.

[0009] Preferably, the detector is a photoelectric sensor, a CMOS or a CCD photoelectric component.

[0010] Preferably, the monochromatic light source includes a monochromatic LED or a laser light source; and the full-band light source includes a multi-spectrum white light source or a halogen lamp full-band light source.

[0011] The technical effects of solving the technical problem in this application are as follows: Compared with the prior art, the PPG multi-channel spectral diffuse reflectance food roasting degree detection method of the present invention adopts the following steps: S1: select multiple monochromatic light sources or full-band light sources and detectors, and each light source and detector is matched with a homogenizer; use the light source to emit light of the required band, and disperse the light through the light source homogenizer, so that the light is evenly irradiated on the sample to be tested; S2: after the light passes through the sample, it generates emission, and the judgment is made according to the absorptivity and reflectivity of the light of different wavelengths irradiated on the sample; the reflectivity is the ratio of the light intensity that can be reflected back from the sample to the irradiation intensity; S3: the reflected light passes through the detector uniformly The light sheet is homogenized again and then enters the detector; the detector detects the reflectivity of light of different wavelengths, and the reaction on the detector is the reflection intensity; the wavelength with higher reflectivity will get a stronger signal in the detector, and vice versa, the signal of the wavelength with smaller reflectivity detected by the detector will be weaker; S4: The detector converts the detected signal into a digital signal through an IV amplifier or a digital analog acquisition front end, and puts it into a register, and obtains the final agtron value through the algorithm of the main control circuit board. By collecting multi-channel spectral data reflected from the food surface and combining appropriate data processing and algorithms, real-time detection and quantitative analysis of the baking degree of food can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic flow chart of the detection principle of a PPG multi-channel spectral diffuse reflectance food roasting degree detection method of the present invention. DETAILED DESCRIPTION

[0013] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0014] See also Figure 1 The present invention provides a PPG multi-channel spectral diffuse reflectance food roasting degree detection method 1, comprising the following steps: S1: Select multiple monochromatic light sources or full-band light sources and detectors, and each light source and detector is matched with a homogenizer; use the light source to emit light of the required band, and disperse the light through the light source homogenizer, so that the light is evenly irradiated on the sample to be tested; S2: The light is emitted after passing through the sample, and the judgment is made based on the absorptivity and reflectivity of light of different wavelengths irradiated on the sample; the reflectivity is the ratio of the intensity of light reflected back from the sample to the intensity of irradiation; S3: The reflected light is homogenized again by the detector homogenizer, and then enters the detector; the detector detects the reflectivity of light of different wavelengths, and the reflection reflected on the detector is the reflection intensity; the higher the reflectivity of the wavelength, the stronger the signal can be obtained in the detector, and vice versa, the signal of the wavelength with a smaller reflectivity detected by the detector is weaker; S4: The detector converts the detected signal into a digital signal through an IV amplifier or a digital analog acquisition front end, puts it into a register, and obtains the final agtron value through the algorithm of the main control circuit board.

[0015] The present application adopts the following steps: S1: select multiple monochromatic light sources or full-band light sources and detectors, and each light source and detector is matched with a homogenizer; use the light source to emit light of the required band, and disperse the light through the light source homogenizer, so that the light is evenly irradiated on the sample to be tested; S2: after the light passes through the sample, it generates emission, and the absorption rate and reflectivity of light of different wavelengths irradiated on the sample are judged; the reflectivity is the ratio of the intensity of the light that can be reflected back from the sample to the irradiation intensity; S3: the reflected light is homogenized again through the detector homogenizer, and then enters the detector; The detector detects the reflectivity of light of different wavelengths, and the reflection reflected on the detector is the reflection intensity. The higher the reflectivity of the wavelength, the stronger the signal can be obtained in the detector. Conversely, the signal detected by the detector is weaker when it detects a wavelength with a smaller reflectivity. S4: The detector converts the detected signal into a digital signal through an IV amplifier or a digital analog acquisition front end, puts it into a register, and obtains the final agtron value through the algorithm of the main control circuit board. By collecting multi-channel spectral data reflected from the food surface and combining appropriate data processing and algorithms, real-time detection and quantitative analysis of the degree of food baking can be achieved.

[0016] In the step S4, a multivariate linear regression model method is used to calculate the agtron value; the spectral intensities of five bands are preset, represented as 𝑥1, 𝑥2, 𝑥3, 𝑥4, and 𝑥5, where each 𝑥𝑖 is a predictor, representing the spectral intensity of a specific band; the agtron value is 𝑦; the multivariate linear regression model is expressed as y=β0+β1x1+β2x2+β3x3+β4x4+β5x5+ϵ; where 𝛽0 is the intercept term; β1, β2, β3, β4, and β5 are model parameters, representing the weight of the influence of each band on the agtron value; ϵ is an error term, used to capture variations that cannot be explained by the model.

[0017] In step S4, the least squares method is used to estimate the model parameter β value; the goal is to minimize the residual sum of squares, that is, the sum of squares of the difference between the actual agtron value and the agtron value predicted by the model; the parameter estimation formula is: β=(XT X) -1 X T y; X is the design matrix, containing all spectral intensity data and a constant term column (corresponding to the intercept term); Y is a vector containing all observed agtron values; β^ is a vector of parameter estimates; After obtaining the model parameters, the new spectral data is input into the model to calculate the predicted agtron value.

[0018] The detector is a photoelectric sensor, a CMOS or a CCD photoelectric component.

[0019] The monochromatic light source includes a monochromatic LED or a laser light source; the full-band light source includes a multi-spectrum white light source or a halogen lamp full-band light source.

[0020] The specific implementation steps are: the power supply part provides energy supply for the whole system; In a multi-spectral system, light sources, detectors and other parts are required; the light source part mainly provides the source of the multi-spectrum, and a monochromatic light source can be selected, such as a monochromatic LED, a laser light source, etc., or a multi-spectrum white light source or a full-band light source such as a halogen lamp can be selected; the multi-channel spectrum emphasized in the present invention generally adopts multiple monochromatic lights to realize the light source, or uses a full-band white light source or a halogen lamp, and then, the receiving end of the detector sensor is placed with a narrow-band filter of different bands to filter out extra light and realize monochromatic light detection, and then multiple detectors are combined to form a multi-channel spectrum acquisition system.

[0021] The detector part mainly collects the spectral data reflected from the food surface. These data contain the reflectivity information of the food at different wavelengths. Photoelectric sensors, CMOS, CCD and other photoelectric components can be used; In order to achieve a better diffuse reflection effect, the present invention innovatively uses two sets of light homogenizers to achieve better light homogenization and diffuse reflection effects; Uniform light source: The light homogenizer can convert an uneven light source into a uniform light source, so that the measurement system obtains uniform illumination.

[0022] This is very important for optical measurements that require precision and accuracy, especially in applications such as imaging and spectroscopy. Reduce background noise: A uniform light source can reduce background noise, thereby improving the signal-to-noise ratio of the measurement system. This is essential for measurement tasks that require high sensitivity and high resolution. Eliminate influences: A uniform light sheet can eliminate the influence of external environmental factors on the measurement, such as changes in the position of the light source, light scattering, etc. This helps to ensure the stability and consistency of the measurement results.

[0023] The collected spectral signal often needs an IV amplifier or a digital analog front end to convert it into a voltage signal, which is then converted into a digital signal by an ADC; and the algorithm deployed on the main control board is used to calculate and finally determine the calculated value. Determine the final degree of roasting; PPG technology is usually used for blood oxygen saturation and heart rate monitoring. PPG is an analog acquisition front end that uses visible light + infrared light. We use this technology to measure the degree of coffee roasting. Based on this principle, there are many PPG acquisition front ends on the market, which have the advantages of high precision and low price, and have the functions of light source driving and multi-channel optical signal acquisition.

[0024] The IV amplifier can convert the input current into the output voltage. The principle is to use an operational amplifier and a feedback circuit to achieve this.

[0025] In the specific application principle, we first select multiple monochromatic light sources to emit light of the required wavelength band, and then disperse the light through the light source homogenizer, so that it can be irradiated on the sample more evenly. After the light passes through the sample, there will be emission. Then we judge it according to the absorptivity and reflectivity of the sample with different wavelengths. The reflectivity is the ratio of the intensity of the light reflected back from the sample to the intensity of the irradiation. The reflectivity of the same sample is different for light of different wavelengths. Therefore, after using multiple light sources to irradiate the sample, the reflectivity of light of different wavelengths is different after reflection from the sample. The reflected light is homogenized again by the detector homogenizer, and then enters the detector. The detector detects the reflectivity of light of different wavelengths, and the reflection on the detector is the reflection intensity. The higher the reflectivity of the wavelength, the stronger the signal can be obtained in the detector. On the contrary, the signal of the wavelength with a smaller reflectivity detected by the detector is weaker.

[0026] The detector converts the detected signal into a digital signal through an IV amplifier or acquisition front end, puts it into a register, and uses the main control algorithm to obtain the final agtron value. In general: Use the method of multiple linear regression to calculate the value of agtron. Suppose you have the spectral intensity of five bands, which can be expressed as X1, X2, X3, X4, X5, where each Xi is a predictor, representing the spectral intensity of a specific band. The value of Agtron is y. Your multiple linear regression model can be expressed as: y=β0+β1x1+β2x2+β3x3+β4x4+β5x5+ϵ where β0 is the intercept term.

[0027] β1, β2, β3, β4, and β5 are model parameters, representing the weight of each band's influence on the Agtron value.

[0028] ϵ is the error term, which captures the variation that cannot be explained by the model.

[0029] At the same time, the least squares method can be used to estimate the model parameters (β values) The goal is to minimize the residual sum of squares, that is, the sum of the squares of the difference between the actual Agtron value and the Agtron value predicted by the model. The parameter estimation formula is: β = (X T X) -1 X T y X is the design matrix, which contains all the spectral intensity data and a constant term column (corresponding to the intercept term).

[0030] Y is a vector containing all observed Agtron values.

[0031] β^ is a vector of parameter estimates.

[0032] Once you have the model parameters, you can input new spectral data into the model and calculate the predicted Agtron values.

[0033] The beneficial effects of this application are: The diffuse reflection optical measurement system can reduce background noise, thereby improving the signal-to-noise ratio of the measurement system. The uniform light sheet can eliminate the influence of external environmental factors on the measurement, such as changes in the position of the light source, light scattering, etc. This helps to ensure the stability and consistency of the measurement results; and after integrating the multi-spectral measurement system, the multi-spectral measurement system can measure spectral data of multiple wavelengths at the same time, providing richer information. The multi-spectral system can provide higher accuracy and identification capabilities. By analyzing spectral data at different wavelengths, the system can more accurately identify the characteristics and properties of the target object; the multi-spectral system can improve the robustness of the system by comparing data at multiple wavelengths. Even when some wavelengths are interfered or noisy, the system can still achieve accurate measurement through data at other wavelengths.

[0034] Therefore, the measurement method of diffuse reflection + multi-channel spectroscopy can make the measurement more accurate and stable; Points that need to be protected in this application: 1. Use PPG acquisition method to measure the visible light band and infrared band of food roasting degree; Second, the light source homogenizer + collection homogenizer is used to even out the light source and collect signals, reducing signal noise.

[0035] The present invention proposes a solution, which is to use multi-spectrum, that is, to use multiple colors of light for detection. Optical diffuse reflection detection is a detection method based on optical principles, which uses the diffuse reflection characteristics of the surface of an object to obtain information about the object. In the detection of the degree of baking of food, optical diffuse reflection detection can be used to obtain spectral reflection information of the food surface in different wavelength ranges, thereby realizing the detection of the degree of baking of the food. This method is expected to solve the problems of high cost and large error in traditional methods, thereby realizing more accurate and economical detection of the degree of baking of food.

[0036] The technical effects of solving the technical problem in this application are as follows: Compared with the prior art, the PPG multi-channel spectral diffuse reflectance food roasting degree detection method 1 of the present invention adopts the following steps: S1: select multiple monochromatic light sources or full-band light sources and detectors, and each light source and detector is matched with a homogenizer; use the light source to emit light of the required band, and disperse the light through the light source homogenizer, so that the light is evenly irradiated on the sample to be tested; S2: after the light passes through the sample, it generates emission, and the judgment is made according to the absorptivity and reflectivity of the light of different wavelengths irradiated on the sample; the reflectivity is the ratio of the light intensity that can be reflected back from the sample to the irradiation intensity; S3: the reflected light passes through the detector uniformly The light sheet is homogenized again and then enters the detector; the detector detects the reflectivity of light of different wavelengths, and the reaction on the detector is the reflection intensity; the wavelength with higher reflectivity will get a stronger signal in the detector, and vice versa, the signal of the wavelength with smaller reflectivity detected by the detector will be weaker; S4: The detector converts the detected signal into a digital signal through an IV amplifier or a digital analog acquisition front end, and puts it into a register, and obtains the final agtron value through the algorithm of the main control circuit board. By collecting multi-channel spectral data reflected from the food surface and combining appropriate data processing and algorithms, real-time detection and quantitative analysis of the baking degree of food can be achieved.

[0037] The above-described embodiments of the present invention do not constitute a limitation on the protection scope of the present invention. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A PPG multi-channel spectral diffuse reflectance food roasting degree detection method, characterized in that: The following steps are included: S1: Select multiple monochromatic light sources or full-band light sources and detectors, and each light source and detector is matched with a light homogenizer; use the light source to emit light of the required band, and disperse the light through the light homogenizer, so that the light is evenly irradiated on the sample to be tested; S2: The light is emitted after passing through the sample, and the judgment is made based on the absorptivity and reflectivity of light of different wavelengths irradiated on the sample; the reflectivity is the ratio of the intensity of light reflected back from the sample to the intensity of irradiation; S3: The reflected light is homogenized again by the detector homogenizer, and then enters the detector; the detector detects the reflectivity of light of different wavelengths, and the reflection reflected on the detector is the reflection intensity; the higher the reflectivity of the wavelength, the stronger the signal can be obtained in the detector, and vice versa, the signal of the wavelength with a smaller reflectivity detected by the detector is weaker; S4: The detector converts the detected signal into a digital signal through an IV amplifier or a digital analog acquisition front end, puts it into a register, and obtains the final agtron value through the algorithm of the main control circuit board.

2. A PPG multi-channel spectral diffuse reflectance food roasting degree detection method as claimed in claim 1, characterized in that: In the step S4, a multivariate linear regression model method is used to calculate the agtron value; the spectral intensities of five bands are preset, represented as 𝑥1, 𝑥2, 𝑥3, 𝑥4, and 𝑥5, where each 𝑥𝑖 is a predictor, representing the spectral intensity of a specific band; the agtron value is 𝑦; the multivariate linear regression model is expressed as y=β0+β1x1+β2x2+β3x3+β4x4+β5x5+ϵ; where 𝛽0 is the intercept term; β1, β2, β3, β4, and β5 are model parameters, representing the weight of the influence of each band on the agtron value; ϵ is an error term, used to capture variations that cannot be explained by the model.

3. A PPG multi-channel spectral diffuse reflectance food roasting degree detection method as claimed in claim 2, characterized in that: In step S4, the least squares method is used to estimate the model parameter β value; the goal is to minimize the residual sum of squares, that is, the sum of squares of the difference between the actual agtron value and the agtron value predicted by the model; the parameter estimation formula is: β=(X T X) -1 X T y; X is the design matrix, containing all spectral intensity data and a constant term column; Y is a vector containing all observed agtron values; β^ is a vector of parameter estimates; after obtaining the model parameters, the new spectral data is input into the model to calculate the predicted agtron value.

4. A PPG multi-channel spectral diffuse reflectance food roasting degree detection method according to any one of claims 1 to 3, characterized in that: The detector is a photoelectric sensor, a CMOS or a CCD photoelectric component.

5. A PPG multi-channel spectral diffuse reflectance food roasting degree detection method as claimed in any one of claims 1 to 3, characterized in that: The monochromatic light source includes a monochromatic LED or a laser light source; the full-band light source includes a multi-spectrum white light source or a halogen lamp full-band light source.

Images