Rapid non-invasive detection method for active ingredients in drying process of colored potatoes

Abstract

The invention discloses a rapid non-invasive detection method for active ingredients in a drying process of colored potatoes, and belongs to the technical field of agricultural product detection. The method comprises the following steps: firstly, acquiring colored potato chip samples which are treated by different drying processes and are at different drying degrees; secondly, acquiring an absorption coefficient, a reduced scattering coefficient and a fluorescence intensity spectrum of the colored potato slices, and performing physical and chemical testing on two active components, namely anthocyanin and total polyphenol, of the colored potato slices; carrying out correlation analysis on the multi-element optical characteristics and each physicochemical index; and finally, based on multi-element optical feature fusion under different spectrum pretreatment, establishing an anthocyanin and total polyphenol quantitative prediction model, and determining an optimal model. According to the method for detecting the active ingredients of the colored potatoes in the drying process, a novel rapid detection method can be provided for ingredients such as anthocyanin and total polyphenol of the potatoes under different drying processes and drying time nodes, sample pretreatment is not needed, and chemical reagents are not consumed.

Description

Technical Field

[0001] This invention belongs to the field of agricultural product testing technology, specifically relating to a rapid, non-invasive method for detecting active ingredients during the drying process of colored potatoes. Background Technology

[0002] Potatoes, as a high-moisture agricultural product, are highly susceptible to sprouting and rotting during storage and transportation, leading to quality deterioration. Drying, by reducing the moisture content of food to a suitable level, thereby preventing microbial growth and extending shelf life, is the main processing method in the potato industry chain. However, during the drying and dehydration process, the specific drying methods and conditions can affect the nutritional components and sensory characteristics of potatoes. Especially for agricultural products rich in heat-sensitive substances such as anthocyanins and other polyphenols, dynamic monitoring and control of the drying process, and modification of drying conditions based on product quality assessment, are essential to obtain dehydrated products with high nutritional value.

[0003] Currently, optical detection technology has been applied in the quality analysis of various agricultural products. A prime example is a rapid quality detection method for potatoes during low-temperature storage and subsequent room-temperature sugar reversion (publication number CN120352362A). This technology focuses on the low-temperature storage (4℃) and subsequent room-temperature sugar reversion (25℃) process of potatoes, monitoring the slow evolution of quality during storage. It uses four macroscopic basic quality indicators—starch, soluble solids (SSC), moisture, and hardness—as core detection targets, reflecting storage stability and processing suitability. This technology provides a solution for monitoring storage quality, and the dual integrating sphere measurement technology it employs is one of the known techniques in this field.

[0004] However, the above-mentioned existing technologies have obvious limitations, making them unsuitable for the detection of active ingredients during the drying process: (1) The above schemes are designed for slow and relatively stable storage environments. The drying process is a violent and dynamic dehydration and thermal processing process, in which the moisture inside the material evaporates rapidly, the temperature changes drastically, and the microstructure is fundamentally reshaped. This rapid evolution of the physicochemical state leads to an essential difference between the optical property change law and the storage process, and directly applying the existing model will fail due to signal distortion. (2) Existing technologies focus on macroscopic basic indicators such as starch and moisture, and cannot effectively reflect the degradation dynamics of high-value, heat-sensitive trace active ingredients such as anthocyanins and total polyphenols during the drying process. These active ingredients are extremely sensitive to heat and light, and their retention rate is the key to measuring the quality of drying processing, but existing technologies lack corresponding detection objects and models. (3) Existing technologies rely only on absorption and scattering coefficients and fail to introduce a more direct and sensitive detection dimension for active ingredients such as anthocyanins and polyphenols. During the violent changes in such matrices during drying, serious spectral interference will occur. The single optical technology has insufficient anti-interference ability and it is difficult to achieve accurate quantification of active ingredients. Furthermore, traditional methods for evaluating and controlling the dehydration process have primarily relied on destructive, cumbersome, and time-consuming chemical methods. In contrast, optically based detection methods have demonstrated significant potential in non-destructively monitoring changes in quality parameters during the drying process of agricultural products. Targeting the unique fluorescence characteristics of phenolic active substances such as anthocyanins and chlorogenic acid in plant-derived agricultural products, fluorescence spectroscopy has shown significant advantages in specificity and detection efficiency in the analysis of active ingredients in agricultural products by specifically detecting the fluorescence signals of target analytes. However, agricultural products contain multiple absorber groups and chromophores, and the complex matrix can easily cause internal filtering effects on the fluorescence signals of the target analytes, leading to a non-linear relationship between fluorescence intensity and fluorophore concentration. Therefore, a single detection technique is insufficient for effective detection and efficient prediction of the target analytes.

[0005] Overcoming the limitations of existing technologies, which are restricted to storage scenarios and macroscopic indicators, and designing a new method specifically for the drastic processing of colored potatoes—drying—to achieve rapid, high-precision, and non-invasive online detection of heat-sensitive active ingredients such as anthocyanins and total polyphenols, is a problem to be solved in this field. Summary of the Invention

[0006] The purpose of this invention is to enable precise, rapid, and non-invasive detection of the active components (total polyphenols and anthocyanins) of different varieties of colored potatoes at different time points during hot air drying and microwave drying. A rapid, non-invasive method for detecting active components during the drying process of colored potatoes is proposed. This method belongs to the field of agricultural product testing technology and provides a new technology for detecting active substances during potato drying processing.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A rapid, non-invasive method for detecting active ingredients during the drying process of colored potatoes, characterized by comprising: (1) Obtain samples of colored potato chips that have been processed by different drying processes and are at different degrees of dryness; (2) The absorption coefficient of potato slices in the 400-1700 nm wavelength range was measured using a dual integrating sphere system. m a With reduced scattering coefficient with me s Simultaneously, a laser-induced fluorescence system was used to measure the fluorescence intensity spectrum of potato slices in the 400-1000 nm wavelength band using a 375 nm wavelength laser as the excitation light. F em To obtain multiple optical characteristics; (3) Physicochemical tests were conducted on potato slices to obtain physicochemical indicators for anthocyanins and total polyphenols; (4) Based on the correlation between the multi-dimensional optical characteristics obtained in step (2) and the anthocyanin content and total polyphenol content measured in step (3), determine the characteristic optical bands that are significantly related to the anthocyanin content and total polyphenol content. (5) Based on the optical data in the characteristic optical band determined in step (3), establish a quantitative prediction model for the content of anthocyanins and total polyphenols, which is used for rapid and non-invasive detection of the active ingredients during the drying process.

[0008] Preferably, the drying process in step (1) includes hot air drying and microwave drying; The hot air drying was carried out at temperatures of 60℃, 70℃, 80℃, and 90℃, and samples were taken at 0, 30, 60, 90, 120, 150, 180, and 210 minutes of drying, respectively. The microwave drying was carried out at power levels of 400W, 480W, 560W, and 640W, and samples were taken at drying times corresponding to 0-26 minutes by adjusting the transmission frequency.

[0009] Two drying methods (hot air drying and microwave drying) and four drying conditions were used, resulting in eight treatment combinations. Each treatment combination had eight sampling time points, with five samples taken each time. A total of 640 valid samples were measured in this experiment: 2 varieties × 2 drying methods × 4 drying conditions × 8 sampling points × 5 replicates.

[0010] Preferably, before drying, the potato sample is washed and peeled, and sliced ​​into thin slices with a thickness of 3.5 ± 0.2 mm using a slicer. The sliced ​​sample is blanched in boiling water for 2 minutes to inhibit polyphenol oxidase activity. The blanched sample is immediately placed in deionized water, protected from light and allowed to stand at room temperature. After removing excess water from the surface with a clean absorbent cloth, it is weighed using an electronic balance.

[0011] Preferably, step (4) includes: Calculate the absorption coefficient and reduced scattering coefficient in the 400-1700 nm band, and the correlation coefficients between the spectral data at each wavelength point of the fluorescence intensity spectrum in the 400-1000 nm band and the anthocyanin content and total polyphenol content. The continuous band with the highest absolute value of the correlation coefficient was selected as the characteristic optical band that was significantly correlated with the anthocyanin content and total polyphenol content.

[0012] Preferably, the characteristic optical bands determined in step (4) include: fluorescence intensity spectrum and absorption coefficient in the 380-780 nm band, absorption coefficient in the 1100-1670 nm band, and reduced scattering coefficient in the 1100-1400 nm band.

[0013] Preferably, step (3) further includes determining the moisture content and color of the potato slices, and the moisture content, color, multivariate optical characteristics, anthocyanin content, and total polyphenol content are used in a correlation analysis.

[0014] Specifically, the moisture content was determined according to the direct drying method in GB 5009.3-2016; The color of the sample was measured using a benchtop spectrophotometer, and the total color difference ( DE ) and browning degree ( BI Calculate using the following formulas respectively:

[0015]

[0016]

[0017] In the formula, L * , a * , b * These are the color indices of dried potato slices; L * 0 , a * 0 , b* 0 The color index is for sliced ​​fresh potatoes.

[0018] Preferably, during the physicochemical testing process, the anthocyanin content is determined by the pH differential method, and the total polyphenol content is determined by the Folin-Ciocalteu colorimetric method.

[0019] Specifically, the anthocyanin content of two varieties of colored potatoes was determined using the pH differential method. First, potato slices that had undergone optical property testing were chopped, and 0.3 g of the chopped sample was accurately weighed and added to 3 ml of anthocyanin extraction solution (1 mol / L hydrochloric acid: methanol = 4:96). The mixture was then ultrasonically extracted at 40 ℃ in the dark for 30 minutes. Then, take two 1ml portions of the extract supernatant and place them in 7ml test tubes. Add 5ml of HCl-KCl buffer (pH 1.0) and HAc-NaAc buffer (pH 4.5) respectively. After equilibration in the dark for 20 minutes, measure the absorbance at 530 nm and 700 nm using a UV spectrophotometer. The formula for calculating anthocyanin content is:

[0020] in, A —Absorbance, A = (A 530nm -A 700 nm )PH1.0 - (A 530nm -A 700 nm pH 4.5; M —The relative molecular mass of cyanidin-3-O-glucoside is 449.2 g / mol; DF —Dilution factor of the test solution; V —Volume of anthocyanin extract, ml; —Molar extinction coefficient of cyanidin-3-O-glucoside, 26900 L / (mol·cm); m—sample mass, g; L —Optical path length, 1 cm; w t — Moisture content of the sample on a wet basis, % The determination of total polyphenol content in colored potatoes was based on the Folin-Ciocalteu colorimetric method with appropriate modifications. First, 0.3 g of chopped potato tissue was accurately weighed and added to 1.5 ml of methanol solution. After sonication in the dark for 2 hours, the mixture was centrifuged at 12000 rpm for 15 minutes. Then take 1 ml of the supernatant and add 1 ml of fluorophenol colorimetric reagent. Shake well, let stand, then add 8 ml of sodium carbonate solution and 10 ml of deionized water. After standing in the dark for two hours, use a UV spectrophotometer to measure the absorbance at 765 nm. A standard curve was established using gallic acid. 5 mg of gallic acid was diluted to 100 ml with deionized water to prepare a 50 μg / ml standard solution. 0, 0.2, 0.4, 0.6, and 0.8 ml of the standard solution were added to 1.0, 0.8, 0.6, 0.4, 0.2, and 0 ml of deionized water, respectively, and then simultaneously measured according to the sample processing procedure. The total polyphenol content of the sample was calculated by substituting it into the standard curve.

[0021] Preferably, the process of establishing and validating the quantitative prediction model includes: 1) Use Savitzky-Golay smoothing pairs m a , with me s and F em The spectrum was smoothed, and principal component analysis was used to analyze the spectrum in the 380-780 nm band. F em and m a-VIS 1100-1670 nm band m a-NIR and the 1100-1400 nm band with me s-NIR Analysis was performed, and the top 7 principal components of various spectral bands were extracted; among them, m a-VIS Indicates the wavelength range of 380-780 nm m a , m a-NIR Indicates the wavelength range of 1100-1670 nm m a , with me s-NIR Indicates the wavelength range of 1100-1400 nm with me s ; 2) Based on spectral compression without PCA F em , m a-VIS , m a-NIR , with me s-NIR After normalizing and fusing different spectra, LSTM networks were used to establish prediction models for anthocyanin and total polyphenol content, respectively. 3) Based on F em , m a-VIS , ma-NIR , with me s-NIR The first 7 principal components were normalized and fused after different spectra were analyzed, and LSTM network was used to establish prediction models for anthocyanin and total polyphenol content, respectively. 4) Compare the prediction effects of the various models obtained in steps 2) and 3) to determine the optimal prediction model for anthocyanins and total polyphenols.

[0022] The preferred optimal prediction model for anthocyanin and total polyphenol content is: F em , m a-VIS , m a-NIR ,and with me s-NIR The LSTM model after data fusion, where F em , m a-VIS , m a-NIR ,and with me s-NIR Principal component analysis was not performed.

[0023] The beneficial effects of this invention are: This invention provides a rapid, non-invasive method for detecting active ingredients during the drying process of colored potatoes. It addresses the current lack of analysis on the dynamic changes in optical properties and major physicochemical qualities during potato drying, the lack of determination of differences in optical properties and physicochemical properties among different varieties of colored potatoes under different drying methods and processes, and the lack of detection technology for anthocyanins and polyphenols, two active ingredients in the potato drying process. This method analyzes the multi-dimensional optical characteristics of different varieties of colored potatoes at different drying degrees under both hot air and microwave drying methods. F em , m a , with me s The method measures and characterizes the main physicochemical indicators and microstructure, conducts correlation analysis between optical property parameters and physicochemical quality, determines the fluorescence interference mechanism, establishes a quantitative prediction model based on multivariate optical feature fusion, and realizes rapid detection of active ingredients (total polyphenols, anthocyanins) during the drying process of colored potatoes. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0025] Figure 1This is a flowchart of a rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes, as proposed in this invention.

[0026] Figure 2 These are actual photos of "Red Beauty" potato slices with different drying methods (hot air and microwave) and different degrees of drying.

[0027] Figure 3 The curves show the dynamic changes in moisture content of (a) "Red Beauty" and (b) "Black King Kong" under different drying times with hot air drying; and the curves show the dynamic changes in moisture content of (c) "Red Beauty" and (d) "Black King Kong" under different drying frequencies with microwave drying.

[0028] Figure 4 The curves show the dynamic changes in anthocyanin content of (a) "Red Beauty" and (b) "Black King Kong" under different drying times with hot air drying; and the curves show the dynamic changes in anthocyanin content of (c) "Red Beauty" and (d) "Black King Kong" under different drying frequencies with microwave drying.

[0029] Figure 5 The curves show the dynamic changes in polyphenol content of (a) "Red Beauty" and (b) "Black King Kong" under different drying times with hot air drying; and the curves show the dynamic changes in polyphenol content of (c) "Red Beauty" and (d) "Black King Kong" under different drying frequencies with microwave drying.

[0030] Figure 6 The two products are (a) "Red Beauty" and (b) "Black Diamond" dried with hot air for different drying times. DE Dynamic change curves; microwave drying of (c) "Red Beauty" and (d) "Black King Kong" at different drying frequencies DE Dynamic change curves; hot air drying of (e) "Red Beauty" and (f) "Black King Kong" at different drying times BI Dynamic change curves; microwave drying of (g) "Red Beauty" and (h) "Black King Kong" at different drying frequencies BI Dynamically changing curve.

[0031] Figure 7 These are microscopic images of potato slices (a), slice cross-sections, and fresh samples and potato powder under two different drying methods.

[0032] Figure 8 The two products are (a) "Red Beauty" and (b) "Black Diamond" dried with hot air for different drying times. m a Spectroscopy; Microwave drying of (c) "Red Beauty" and (d) "Black King Kong" at different drying frequencies m a spectrum.

[0033] Figure 9The two products are (a) "Red Beauty" and (b) "Black Diamond" dried with hot air for different drying times. with me s Spectroscopy; Microwave drying of (c) "Red Beauty" and (d) "Black King Kong" at different drying frequencies with me s spectrum.

[0034] Figure 10 The two products are (a) "Red Beauty" and (b) "Black Diamond" dried with hot air for different drying times. F em Spectroscopy; Microwave drying of (c) "Red Beauty" and (d) "Black King Kong" at different drying frequencies F em spectrum.

[0035] Figure 11 The drying processes of (a) "Red Beauty" and (b) "Black King Kong" potatoes during hot air drying, and (c) "Red Beauty" and (d) "Black King Kong" potatoes during microwave drying. m a-540 nm , with me s-1200 nm as well as F 450 nm Thermograph of correlation with physicochemical properties.

[0036] Figure 12 The samples are (a) "Red Beauty", (b) "Black King Kong", and (c) all samples during hot air drying and (d) "Red Beauty", (e) "Black King Kong", and (f) all samples during microwave drying. m a The full-band correlation coefficient spectrum of total polyphenols, anthocyanins and moisture content.

[0037] Figure 13 The samples are (a) "Red Beauty", (b) "Black King Kong", and (c) all samples during hot air drying and (d) "Red Beauty", (e) "Black King Kong", and (f) all samples during microwave drying. with me s A spectrum of correlation coefficients across the entire spectrum with total polyphenols, anthocyanins, and moisture content.

[0038] Figure 14 The samples are (a) "Red Beauty", (b) "Black King Kong", and (c) all samples during hot air drying and (d) "Red Beauty", (e) "Black King Kong", and (f) all samples during microwave drying. F em A spectrum of correlation coefficients across the entire spectrum with total polyphenols, anthocyanins, and moisture content. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] This invention proposes a rapid, non-invasive method for detecting active ingredients during the drying process of colored potatoes, such as... Figure 1 As shown, its implementation process is as follows: (1) Drying process and sample preparation In this embodiment, two types of colored potatoes, "Red Beauty" and "Black King Kong," were harvested from the field. Potatoes with no obvious surface damage, intact shape, no rot, and no sprouting were selected. Before drying, potatoes with no obvious surface damage, intact shape, no rot, and no sprouting were selected, washed, peeled, and sliced ​​into thin slices with a thickness of 3.5 ± 0.2 mm using a slicer. To avoid browning of the samples due to high temperature during the drying process, the sliced ​​samples were blanched in boiling water for 2 minutes to inhibit polyphenol oxidase activity, thereby reducing nutrient loss and improving quality. The blanched samples were immediately placed in deionized water, protected from light, and allowed to stand at room temperature. After that, excess surface moisture was absorbed with a clean absorbent cloth, and the samples were weighed using an electronic balance (accuracy 0.01 g) before drying.

[0041] Two typical drying methods, hot air drying and microwave drying, were used for drying. Hot air drying employed a forced-air drying oven with four temperature gradients of 60, 70, 80, and 90°C. Samples were taken at 0, 30, 60, 90, 120, 150, 180, and 210 minutes of drying. To prevent shrinkage and deformation of potato slices due to water loss during hot air drying, which could affect the accuracy of subsequent spectral acquisition experiments, pre-treated slices were laid flat on a mesh sample tray lined with filter paper before hot air drying. The tray and filter paper of the same specification were then placed on top to apply uniform pressure and ensure morphological stability during drying. Microwave drying used a continuous tunnel microwave sterilization line with four power gradients of 400, 480, 560, and 640 W. Transmission frequencies of 0, 7, 6, 5, 4, 3, 2, and 1 Hz were set, corresponding to drying times of approximately 0, 3, 4, 5, 6, 8, 13, and 26 minutes. Samples were taken at different transmission frequencies.

[0042] Two drying methods (hot air drying and microwave drying) and four drying conditions were used, resulting in eight treatment combinations. Each treatment combination had eight sampling time points, with five slices sampled each time. A total of 640 valid samples were measured: 2 varieties × 2 drying methods × 4 drying conditions × 8 sampling points × 5 replicates. The photographs of the "Red Beauty" potato slices obtained under different drying methods and degrees of dryness are shown below. Figure 2 As shown.

[0043] (2) Physicochemical index characterization The moisture content was determined according to the direct drying method in GB 5009.3-2016 "National Food Safety Standard - Determination of Moisture in Food".

[0044] Figure 3 (a) and (b) show the dynamic changes in moisture content of two colored potato varieties, "Red Beauty" and "Black King Kong," during hot air drying, respectively. It can be observed that both potato varieties exhibit a relatively slow moisture decrease trend in the initial stage of hot air drying (0-60 minutes), when the material temperature rises from room temperature to the set drying temperature, and the absorbed heat energy is mainly used to heat the material. As the drying time increases, at four different drying temperatures of 60, 70, 80, and 90℃, the rapid dehydration stage begins at 60, 90, 120, and 150 minutes, respectively. However, the drying rate slows down when the moisture content decreases to 10% in the later stages of drying. Previous studies have shown that, based on the hardness characteristics of the samples during the drying stages, the hot air drying process of potatoes can be divided into two stages: a softening stage and a hardening stage. The softening stage occurs in the initial stage of drying, when free water on the sample surface begins to evaporate rapidly, and the cell structure is not yet completely destroyed. In the hardening stage in the later stage of drying, water evaporation is mainly bound water, and the cell walls or intercellular structures collapse, causing the sample to harden. In this embodiment, the rate of moisture loss in the initial drying stage was slower than reported in the literature. This may be because both the upper and lower surfaces of the potato slices were covered with filter paper, preventing the hot air from directly contacting the sample surface. However, the filter paper has good air permeability and does not completely prevent moisture loss. Compared to studies that directly expose the potatoes to hot air, it only reduced the rate of moisture evaporation. A clear transition between the softening and hardening stages could still be observed in both types of colored potatoes, and the transition time point advanced with increasing drying temperature.

[0045] Figure 3(c) and (d) show the dynamic moisture content changes of "Red Beauty" and "Black King Kong" potato slices during microwave drying, respectively. Based on the trend of moisture content changes, under drying power conditions of 400 W and 480 W, a rapid increase in the rate of moisture content decrease can be observed at drying frequencies of 3 Hz and 4 Hz. Potato slices dried at 560 W and 640 W maintained a high dehydration rate throughout the drying process, possibly due to the higher drying power. However, the "Black King Kong" sample showed an abnormal decrease in moisture content at a drying power of 640 W and a drying frequency of 4 Hz, which may be related to the drying method and sampling method. During microwave drying, local overheating and edge effects can cause non-uniform drying, resulting in a non-uniform distribution of moisture content. Furthermore, although microwave drying technology offers a fast drying rate and produces a smooth surface after drying, it is difficult to accurately control the temperature and humidity during drying, thus affecting the color difference and moisture content of the product.

[0046] Food stability depends on its moisture content. Only mobile water molecules can be utilized by microorganisms and enzymes. Therefore, the target moisture content for drying is generally set below 10%. To achieve this target moisture content, hot air drying takes 150-210 minutes, while microwave drying only takes 13-28 minutes, reducing drying time by about 86% and greatly improving drying efficiency.

[0047] Since anthocyanins exist in different forms at different pH values, the pH differential method was used to determine the anthocyanin content of two varieties of colored potatoes. Potato slices of different drying degrees were chopped, and 0.3 g of the chopped sample was accurately weighed and added to 3 mL of anthocyanin extraction solution (1 mol / L hydrochloric acid: methanol = 4:96). The mixture was ultrasonically extracted at 40℃ in the dark for 30 minutes. Two 1 mL aliquots of the extraction supernatant were placed in 7 mL test tubes, and 5 mL of HCl-KCl buffer (pH = 1.0) and HAc-NaAc buffer (pH = 4.5) were added respectively. After equilibration in the dark for 20 minutes, the absorbance was measured at 530 nm and 700 nm using a UV spectrophotometer. The anthocyanin content was calculated using the following formula:

[0048] A —Absorbance, A = ( A 530 nm - A 700 nm ) PH 1.0 - ( A 530 nm - A 700 nm ) PH 4.5 ; M —The relative molecular mass of cyanidin-3-O-glucoside is 449.2 g / mol; DF —Dilution factor of the test solution; V —Volume of anthocyanin extract, mL; —Molar extinction coefficient of cyanidin-3-O-glucoside, 26900 L / (mol·cm); m —Sample mass, g; L —Optical path length, 1 cm; w t — Moisture content of the sample on a wet basis, %.

[0049] Figure 4 As shown in this embodiment, the "Red Beauty" and "Black King Kong" samples contained 143.71 mg / 100g and 428.86 mg / 100g of anthocyanins respectively after blanching. This indicates that the anthocyanin content in purple-skinned and purple-fleshed potatoes is significantly higher than that in red-skinned and red-fleshed potatoes.

[0050] Anthocyanins, as heat-sensitive substances, are easily degraded under high temperatures. The anthocyanin content in both types of colored potatoes showed a decreasing trend during both hot air drying and microwave drying processes. During hot air drying, the degradation of anthocyanin content can be divided into three stages. In the initial drying stage, the anthocyanin content decreases slowly. During this process, the material is in a slow heating phase, and the initial evaporation of surface moisture absorbs a large amount of heat, creating a high-humidity microenvironment that hinders oxygen diffusion, resulting in a low anthocyanin degradation rate. Furthermore, since anthocyanins are water-soluble pigments, high water content has a certain protective effect on them; therefore, the anthocyanin content decreases slowly as moisture evaporates in the initial drying stage. In the middle drying stage, under four drying temperatures (60, 70, 80, and 90℃), the anthocyanin content shows a sharp decrease after 150, 120, 90, and 60 minutes of drying, respectively, with a decrease of up to 55.34%. At this stage, the potato sample is in the initial hardening phase, rapidly losing water and transitioning from a "rubbery" to a "glassy" state. This rapid evaporation of water is accompanied by the rapid degradation of the water-soluble pigment anthocyanin. Simultaneously, the reduction in moisture causes the rupture of cell walls and vacuolar membranes, leading to cell structure collapse and the release of large amounts of previously protected anthocyanins, exposing them to high temperatures and oxygen. Furthermore, at high temperatures (90°C), water evaporation is faster, reaching the critical moisture content earlier, while at low temperatures (60°C), moisture migration is slower, delaying the critical point for rapid anthocyanin degradation to 150 minutes, consistent with the decreasing trend in moisture content. Towards the end of drying, the decreasing trend in anthocyanin content slows again, and the content tends to stabilize. At this stage, the moisture content in the material decreases to below 15%, placing the material in a "glassy" state. This extremely low moisture content may cause some anthocyanins to bind tightly to cell residues, making them difficult to degrade. Simultaneously, the remaining anthocyanins may transform into derivatives with higher thermal stability.

[0051] During microwave drying, the anthocyanin content showed a significant decrease in the initial sampling stage (drying frequency 7 Hz, drying time approximately 3 minutes), with a degradation rate of about 21.71%. Afterward, it exhibited a step-like, stable decline. This is because microwave drying is highly efficient; in the initial stage, microwaves directly act on the water molecules inside the sample, causing rapid evaporation of water and a surge in vapor pressure. This leads to rapid rupture of cell walls and vacuolar membranes in the early stages of drying, resulting in the instantaneous release of anthocyanins and accelerated thermal oxidation and degradation. Subsequently, the anthocyanin content decreases as the moisture content continues to decrease steadily. Furthermore, the higher the microwave intensity, the faster the water evaporation and the greater the amount of anthocyanin degradation.

[0052] The anthocyanin degradation results under both drying methods indicate that the degradation of anthocyanins is a non-spontaneous endothermic reaction, and the degradation amount increases with decreasing moisture content. During hot air drying, the anthocyanin loss rate increased from 82.86% (60°C) to 87.57% (90°C) with increasing drying temperature, following the principles of first-order reaction kinetics. Simultaneously, during microwave drying, the anthocyanin loss rate also increased from 81.25% to 88.47% with increasing drying power from 400 W to 640 W.

[0053] The determination of total polyphenol content in colored potatoes was performed using the Folin-Ciocalteu colorimetric method with appropriate modifications. 0.3 g of chopped potato tissue was accurately weighed and added to 1.5 mL of methanol solution. After sonication in the dark for 2 hours, the mixture was centrifuged at 12000 rpm for 15 minutes. 1 mL of the supernatant was collected and 1 mL of Folin-Ciocalteu colorimetric reagent was added. After shaking well, 8 mL of sodium carbonate solution and 10 mL of deionized water were added. After standing in the dark for two hours, the absorbance was measured at 765 nm using a UV spectrophotometer.

[0054] A standard curve was constructed using gallic acid: 5 mg of gallic acid was diluted to 100 mL with deionized water to prepare a 50 μg / mL standard solution. 0, 0.2, 0.4, 0.6, and 0.8 mL of the standard solution were added to 1.0, 0.8, 0.6, 0.4, 0.2, and 0 mL of deionized water, respectively, and then simultaneously measured according to the sample processing procedure. The total polyphenol content of the sample was calculated by substituting the values ​​into the standard curve.

[0055] Figure 5 This study demonstrates the dynamic changes in total polyphenol content in two types of colored potatoes during hot air drying and microwave drying. It shows that increasing drying time and temperature / power significantly leads to polyphenol decomposition, and the effects of both drying methods on total polyphenol content follow a similar trend to their effects on anthocyanin content. This is because the content of flavonoids and total polyphenols in potato tubers shows a significant positive correlation, and anthocyanins are the main flavonoid in colored potatoes. However, the degradation rate of total polyphenols is higher in the later stages of drying, with some sampling points even showing an increase in total polyphenol content. Research indicates that potato slices rich in carbohydrates and polyphenols are prone to Maillard reactions during high-temperature drying. The resulting melanoidins and phenolic substances have some overlapping biological characteristics and often coexist in heat-processed products. Therefore, when using the Folin-Ciocalteu colorimetric method to determine total polyphenol content, this non-enzymatic browning product may partially mask the actual polyphenol loss, leading to an overestimation of the measured total polyphenol content. This phenomenon has been reported in various heat-processed food systems, including fruit and vegetable products and baked grains.

[0056] To investigate the color change of potato slices during the drying process, a benchtop spectrophotometer (CM-5 Konica Minolta) was used to measure the color of the samples. Each sample was measured five times, and the average value was taken. The total color difference (…) DE ) and browning degree ( BI Calculate using the following formulas respectively: In the formula, L * , a * , b * These are the color indices of dried potato slices; L * 0 , a * 0 , b * 0 The color index is for sliced ​​fresh potatoes.

[0057] in X Calculated as follows: like Figure 6 As shown, observations can be made on potato slices of both varieties during hot air and microwave drying processes. DE and BI The color of potato slices increased significantly. Due to the blanching pretreatment, polyphenol oxidase was deactivated to some extent. Therefore, the color change of potato slices during drying was mainly caused by non-enzymatic browning and degradation of pigments.

[0058] During the hot air drying process, DE and BI There was no clear linear relationship with drying temperature; samples dried at 60°C and 80°C showed higher [quality / efficacy] compared to those dried at 90°C. DE as well as BI This is because under high-temperature conditions (80-90°C), the continuous heating in the initial stage of drying (0-90 minutes) causes rapid dehydration of the slice surface, leading to an intensified non-enzymatic browning reaction. DE as well as BI The moisture content rises rapidly; however, as the drying process progresses (after 90 minutes), the surface moisture is rapidly lost, the material surface hardens, and the lower moisture content inhibits the browning reaction from continuing, resulting in... DE as well as BIThe growth rate slowed significantly. Under low-temperature conditions (60-70°C), the drying time was long, the moisture content decreased more slowly, and the browning reaction lasted longer, leading to the continuous accumulation of browning products and ultimately... DE as well as BI The color deterioration level is similar to that achieved with high-temperature treatment.

[0059] In the initial stages of microwave drying (7 Hz and 6 Hz), the sample surface turned white and caused... L * The increase in value is a result of significant sugar leaching, which exposes a whiter surface and increases the brightness value. Furthermore, related studies also suggest that microwave drying leads to a higher level of anthocyanin degradation. Simultaneously, it can be observed that as the microwave drying power increases, the sample's... BI The significant increase is likely due to the longer reaction time and higher reaction power, which cause charring of the sample surface and exacerbate browning.

[0060] The microscopic morphology of potato samples was observed using a scanning electron microscope (SEM, Quanta-200, Netherlands). The experimental subjects included potato slices (surface and cross-section) under two drying methods, whole potato powder, and fresh tubers.

[0061] During sample preparation, samples with the lowest moisture content obtained under hot air drying (210 min) and microwave drying (7 Hz) conditions were selected as dried section samples for observation of their surface and cross-sectional microstructure. The dried sections were then pulverized and passed through a 40-mesh sieve to obtain whole powder samples. Fresh potato tubers (3 × 3 mm) were fixed with 2.5% glutaraldehyde solution at 4℃ for 2 hours and rinsed three times with phosphate buffer (pH = 7.2) for 15 min each time. Then, they were fixed with 1% osmium tetroxide solution at 4℃ for 2 hours and rinsed three times with phosphate buffer (pH = 7.2) for 30 min each time. The cubes were then subjected to gradient dehydration with 50%, 70%, 80%, 90%, and 100% acetone solutions, 30 min for each gradient. After reaching the critical point of drying, these were used as fresh potato tuber samples for testing. After ion sputtering gold sputtering, the three types of samples were placed in the electron microscope sample chamber, vacuum extracted, and the working voltage set to 5 kV for microscopic morphological observation.

[0062] From Figure 7As observed in (a), both drying methods resulted in irregular porous structures on the surface of potato slices, and the effects of different drying methods on the microstructure of potato slices showed significant differences. During hot air drying, the samples gradually formed a porous honeycomb structure as the temperature increased. In contrast, although microwave-dried samples also formed a honeycomb structure, the arrangement of its pores showed significant irregularity, which may be related to the uneven evaporation of moisture during microwave heating. At the same time, due to its stronger dehydration intensity, microwave drying resulted in more pores appearing on the sample surface as the drying power increased.

[0063] (3) Acquisition of multi-dimensional optical properties After drying, potato slices were removed, placed in a glass desiccator, and cooled to room temperature in the dark before their optical properties were measured.

[0064] Transmittance was measured using a dual integrating sphere system. T ) and reflectivity ( R For spectral acquisition, since the anthocyanin and polyphenol content of potatoes is unevenly distributed throughout the potato, in order to reduce measurement errors, multiple points in the outer medulla of the potato were selected as detection points, and the average value was taken.

[0065] Using a micrometer, three points were taken at the edge and center of the potato slice to measure the thickness, and the average value was used as the thickness parameter required for the IAD algorithm.

[0066] Measure the refractive index of potato samples using an Abbe refractometer n The final measurement result was 1.34.

[0067] Based on the measured data using the IAD algorithm... R , T glass slides and potatoes n , d The absorption in the 400-1700 nm range of potato slices was calculated. m a ) and reduced scattering coefficient ( with me s )spectrum.

[0068] In the laser-induced fluorescence system, a 375 nm laser is used as the excitation light to irradiate potato slices. The emitted fluorescence passes through a long-pass filter (transmission range 400-1000 nm) located at the receiving end in reflection mode, and is then analyzed by a spectrometer at the receiving end to obtain the fluorescence emission intensity at 400-1000 nm. F em ) like Figure 8 The image shows "Red Beauty" and "Black Diamond" tea leaves after different drying times using hot air and microwave drying methods. ma Spectroscopy. Both colored potatoes exhibit two distinct absorption peaks in the 320-800 nm range. The characteristic absorption peak at 320-400 nm is mainly caused by the combined effects of phenolic acids and flavonoids in the phenolic compounds; the characteristic absorption peak at 500-700 nm is caused by... β - This is caused by the combined action of carotenoids and anthocyanins, but for potatoes with colored flesh, β Carotenoids are usually masked by darker anthocyanins. Furthermore, due to differences in the types and amounts of anthocyanins in the two colored potatoes, the anthocyanin absorption peak of "Red Beauty" is approximately 520-580 nm, while that of "Black King Kong" shows a slight blue shift to 540-600 nm. After a short blanching heat treatment, the high temperature causes the cell membranes and cell walls to rupture, releasing more pigments and resulting in a stronger characteristic absorption peak. Therefore, compared to fresher potato samples, the heat-treated samples... m a The shape and intensity of the spectrum have both changed.

[0069] Two types of colored potatoes during hot air drying m a In the visible light band, the absorption capacity of anthocyanins and carotenoids showed a slow increase in the early stage of drying, followed by a rapid increase in the later stage. In the early stage of drying, slow evaporation of moisture led to an increase in the relative concentration of pigments such as anthocyanins and carotenoids, and the formation of their degradation products also enhanced the sample's absorption capacity. In the middle and later stages of drying, the non-enzymatic Maillard reaction intensified, and the resulting intermediate products (5-HMF) and melanoidins exhibited absorption characteristics at 420-450 nm, further promoting absorption. m a The increase in [something]. During the microwave drying process, the "Red Beauty" [something]... m a In the visible light band, it also shows a continuous upward trend, but the growth rate is more stable than that of hot air drying, which is consistent with its moisture loss rate. During the early stages of microwave drying, the "Black Diamond" exhibits a similar trend in the visible light band. m a The sustained increase in anthocyanin content is also due to pigment changes and intensified browning reactions during prolonged heating. The subsequent decrease in the later stages of drying may be related to the anthocyanin composition and processing methods. Studies have shown that the chemical structure of major anthocyanins may affect potato stability, and the deterioration of anthocyanins under heat treatment may be related to their morphology, processing conditions, and the interaction between matrix components. Red-fleshed potatoes have the highest content of pelargonidin derivatives, while petuniatin and malvidin are the most abundant anthocyanins in purple potatoes. Therefore, the difference in anthocyanin composition leads to variations in the anthocyanin content of "Red Beauty" and "Black King Kong" potatoes during the drying process. m aThe changes are different. The slow and steady dehydration process of hot air drying may mask this difference, while in microwave drying, where the dehydration process is more intense, the huge changes in internal pressure cause cell rupture, making it easier for less stable pigments to be lost and altering the structure. Furthermore, in the later stages of microwave drying, over-drying leads to a deterioration in sample quality; the Maillard reaction stops due to insufficient substrate, and melanoidin production decreases or even decomposes. In the near-infrared band, m a A weak absorption peak is observed near 980 nm, which is related to the moisture and soluble solids in the sample; while the strong absorption peak in the 1400 nm-1700 nm range is caused by the combined effects of moisture, starch, and sugars. During the drying process, this wavelength range involves multiple complex mechanisms, including moisture reduction, starch restructuring, and the participation of reducing sugars in the Maillard reaction. m a The overall trend shows a downward trend with relatively obvious waveform changes.

[0070] During the hot air drying and microwave drying processes, the "Red Beauty" sample was dried at 540 nm. m a The average increase was 1124.75% and 572.19%; "Black Diamond" showed a higher concentration at 540 nm during both drying processes. m a The average increase was 752.53% and 187.41%, respectively. This indicates that hot air has a stronger effect on the changes in the absorption characteristics of the samples than microwave drying.

[0071] like Figure 9 The image shows "Red Beauty" and "Black Diamond" tea leaves after different drying times using hot air and microwave drying methods. with me s Spectra. It can be seen that potato slice samples under different drying methods... with me s The changes are significant, showing a clear upward trend overall. This is because light scattering depends on changes in the microstructure of the sample tissue, such as its hardness, cells, and intercellular spaces, and the water loss during the drying process leads to strong changes in this microstructure. In both drying methods, with me s The light intensity is generally flat in the 400-900 nm wavelength range, decreasing with increasing wavelength, and generally conforms to the Mie scattering law. However, in the later stages of drying... with me s A peak appears at 540 nm and intensifies with increasing drying time. This is likely due to the enhanced absorption of pigments and Maillard reaction products in this band during the later stages of drying, causing crosstalk to the scattering. Meanwhile, strong absorption of starch, water, and sugars occurs in the near-infrared 1400-1700 nm band, which affects the inversion solution of the IAD algorithm. with mes This caused interference, and the inherent crosstalk drawbacks of the dual integrating sphere led to abnormal scattering peaks in this band. Later in the drying process, as moisture evaporated and nutrients degraded in the sample... m a The decrease is accompanied by with me s The enhancement, with me s The crosstalk effects in this band gradually weaken, and the overall spectrum returns to a smoother state.

[0072] During the hot air drying process, the two types of colored potatoes in the early stage of drying... with me s It shows a gradual upward trend, with an increase of approximately 18.17% per 30 minutes; after entering the later stage of drying, with me s The value shows a significant upward trend, with a maximum increase of 86.18% / 30min (taking "Red Beauty" potato slices dried at 60℃ as an example). This is partly due to the increase in particle density and porosity caused by shrinkage and moisture loss in the potato slices during drying, which in turn leads to... with me s On the one hand, high temperature increases the rate of water removal from tissues, but it also leads to surface hardening. The rate of water loss determines the speed of change in the sample's tissue structure, thus affecting... with me s The rate of change. In the early stage of hot air drying, the sample is in a softening stage, at which time the rate of water loss is relatively slow, and the evaporation of surface free water does not completely destroy the cell structure. The cells still contain a large amount of water, resulting in relatively large cell wall spaces and weaker light scattering. In the later hardening stage of drying, water evaporation is mainly bound water, the intercellular structure begins to collapse, and the sample transforms from a "rubber state" to a "glass state," increasing its hardness. Microwave drying, on the other hand, directly acts on the water molecules inside the sample, exhibiting a high dehydration rate throughout the process. Therefore, "Red Beauty" exhibits a higher dehydration rate during microwave drying. with me s The upward trend was relatively constant, and the rate of increase was positively correlated with the drying power, which is consistent with the rate of moisture loss during microwave drying. However, during the microwave drying process of "Black Diamond," under three higher power (480, 560, 640W) drying conditions, with me s The subsequent decline may be due to the damage and collapse of the tissue structure caused by continuous high-power drying.

[0073] The "Red Beauty" sample was dried in hot air at a wavelength of 1200 nm. with me s The average increase was 470.90%, while in microwave drying... with me s The average increase was 531.48%; the "Black Diamond" at 1200 nm during the hot air drying process... with me s The average increase was 443.66%, while in microwave drying... with me s The average increase was 299.73%. Differences in drying methods led to variations in the samples. with me s The trends of change are different. When hot air is dried, the free water on the surface of the sample begins to evaporate, while the bound water inside only begins to be lost in the later stage of drying. The dried sample has a compact structure and a dense, hardened layer on the surface. When microwave is dried, the moisture inside and outside the sample evaporates at the same time, and the dried sample is loose and porous.

[0074] like Figure 10 The image shows "Red Beauty" and "Black Diamond" tea leaves after different drying times using hot air and microwave drying methods. F em Spectroscopy. Both types of blanched colored potatoes exhibited distinct fluorescence peaks in the 400-500 nm wavelength range. The peak for "Red Beauty" was located at 470 nm, while the peak for "Black King Kong" showed a slight red shift to 465 nm. This fluorescence emission peak is primarily due to chlorogenic acid, a phenolic compound and the main endogenous fluorescent substance in potatoes, which emits characteristic fluorescence around 450 nm. However, unlike previous studies reporting that fresh purple potatoes exhibited the strongest fluorescence peak at 526 nm, the characteristic fluorescence peak of anthocyanins (143.705 mg / 100g) in blanched "Red Beauty" potatoes showed a significant blue shift to 580-620 nm. In contrast, "Black King Kong" potatoes (428.863 mg / 100g), which have a higher anthocyanin content, did not show a distinct anthocyanin characteristic fluorescence peak in this range. This is attributed to interference from the internal filtering effect. The internal filtering effect is a radiative energy transfer phenomenon, which can be divided into a primary effect—the excitation light is absorbed by absorbing groups before reaching the fluorophore, leading to intensity attenuation—and a secondary effect—the emitted fluorescence of the fluorophore is reabsorbed by other components before detection. The synergistic effect of these two effects results in a non-linear relationship between fluorophore concentration and fluorescence intensity, thus masking the characteristic fluorescence signal of anthocyanins. "Black Diamond" at 600 nm... m a (0.324 mm) -1 It is approximately the size of "Red Beauty" (0.064 mm). -1The anthocyanin content is five times higher than that of "Black King Kong" (2793.673 mg / 100g), and the characteristic fluorescence produced by anthocyanins themselves is weaker. This leads to a greater impact of the internal filtering effect on the detection of the characteristic fluorescence of anthocyanins in "Black King Kong". This also explains why the total phenol content of "Black King Kong" (2793.673 mg / 100g) is higher than that of "Red Beauty" (2017.802 mg / 100g), but its fluorescence intensity is much lower than that of "Red Beauty".

[0075] During the drying process, the characteristic fluorescence peak of phenolic substances at 400-500 nm showed a trend of first decreasing and then increasing. This is because in the early stage of drying, other active substances in colored potatoes, such as chlorogenic acid and caffeic acid, are thermally degraded, leading to a decrease in the content of fluorophores and thus a decrease in fluorescence intensity. In the later stage of drying, with the increase of drying time, melanoidins, intermediate and final products of the Maillard reaction, gradually become the dominant fluorescent substances. Melanoidins are a highly heterogeneous mixture with pale yellow and blue-green fluorescence, possessing a large number of double-bond conjugated systems. They emit characteristic fluorescence in the 400-500 nm band, with a maximum emission peak at approximately 440 nm. In the later stage of drying, the accumulation of melanoidins leads to an increase in fluorescence intensity in this band and a redshift of the peak value (approximately 5 nm). In the later stage of microwave drying, the fluorescence initially increased and then decreased. This may be due to the insufficient Maillard reaction substrate within the material caused by the strong drying conditions, resulting in the termination of the reaction and the degradation of melanoidins. The fluorescence peak appearing at 640 nm in the later stage of drying may be related to the oxidation products of anthocyanins.

[0076] (4) Correlation analysis between optical properties and physicochemical indicators In this embodiment, the selection m a-540 nm , with me s-1200 nm as well as F 450 nm As characteristic optical parameters, physicochemical indicators include total polyphenol content, anthocyanin content, moisture content, and colorimetric value (L). * a * b * Correlation analysis of optical properties and physicochemical indicators was conducted. Figure 11 Thermographs showing the correlation between optical properties and physicochemical properties of two colored potato samples under different drying methods (hot air and microwave) are presented.

[0077] like Figure 11 As shown in (a) and (b), during the hot air drying process, the two types of colored potatoes... m a-540 nm , with me s-1200 nm It showed a significant negative correlation with total polyphenols, anthocyanins, and moisture content (correlation coefficient r). < -0.8), this is because the drying process not only leads to a continuous decrease in the content of the three physicochemical substances, but also causes continuous deterioration of heat-sensitive substances and Maillard browning. The synergistic effect of these two factors leads to the sample... m a The number continues to increase. "Red Beauty" F 450 nm The fluorescence properties of "Black King Kong" showed a strong correlation with the three physicochemical indicators (r ≤ -0.65), while the fluorescence properties showed a lower correlation with the physicochemical indicators (0.49 ≤ r ≤ 0.60). This may be because the high anthocyanin content of "Black King Kong" exacerbated the strong absorption, which aggravated the internal filtering effect on the fluorescence detection accuracy, thus reducing the fluorescence detection accuracy. F The correlation between it and physicochemical indicators. Figure 11 Two types of colored potatoes can also be found in it. F 450 nm and m a-540 nm , with me s-1200 nm There is a strong correlation between them (0.62 ≤ r ≤ 0.81), which provides a reference for subsequent methods to improve the model prediction accuracy based on the fusion of multiple optical properties.

[0078] like Figure 11 As shown in (c) and (d), during the microwave drying process, "Red Beauty" m a-540 nm , with me s-1200 nm as well as F 450 nm The negative correlation between total polyphenols, anthocyanins, and moisture content was significant (-0.97 ≤ r ≤ -0.70), while that between "Black King Kong" and... with me s-1200 nm and F 450 nm Although it shows a strong correlation with physicochemical indicators (-0.90 ≤ r ≤ -0.74), its m a-540 nm The correlation with physicochemical indicators weakened significantly (-0.5 < r < 0). This may be due to the "black diamond" material's... m a Although significant changes occurred, unlike the monotonous increasing trend of "Red Beauty," the differences in anthocyanin thermal stability led to its... m a During microwave drying, the temperature initially rises and then falls. This non-monotonic change also contributes to the "Black Diamond's" appearance. m a-540 nm , with me s-1200 nm , F450 nm The weakening of the correlation between them, while the "Red Beauty" F 450 nm and m a-540 nm , with me s-1200 nm There is still a strong correlation between them (0.66 ≤ r ≤ 0.94).

[0079] Figure 12 , 13 14 and 14 represent colored potatoes dried using two different methods, respectively. m a , with me s , F em The correlation curves across the entire wavelength range between the optical properties and total polyphenols, anthocyanins, and moisture content are shown. It can be seen that the three physicochemical indicators exhibit similar correlation curves with the optical properties. This may be because polyphenols such as anthocyanins are water-soluble, and the strong correlation among the three (r ≥ 0.95) is also evident. Figure 11 This was observed in [the context].

[0080] from Figure 12 It can be observed that during the hot air drying process, the wavelength range of 400-900 nm... m a A relatively stable negative correlation was observed with the three physicochemical components (r < -0.75). However, the correlation curve fluctuated and increased after 900 nm, with sharp correlation peaks appearing near 1390 nm and 1538 nm. This is due to the strong absorption characteristics of moisture and sugars in this wavelength range. During microwave drying, the 400-900 nm wavelength range... m a The negative correlation curve with physicochemical properties shows a relatively obvious peak and valley in the characteristic absorption peak region of phenolic substances (400-600 nm), and this is due to the fact that "Black Diamond" is subjected to microwave drying. m a The non-monotonicity of the changes disrupts the collinearity between spectral information, resulting in abnormal fluctuations in the correlation curve in the 600-800 nm band. Compared with hot air drying, microwave drying shows a more rapid increase in the correlation curve between optical properties and physicochemical properties after 900 nm, and broad and flat peaks appear at 1390 nm and 1500 nm.

[0081] like Figure 13 As shown, during the hot air drying process, with me sThe correlation with physicochemical indicators shows a negative correlation across the entire wavelength range, and the correlation curve fluctuates significantly at the characteristic absorption bands of phenolic substances (320-400 nm), carotene, and anthocyanins (500-600 nm), which may be due to crosstalk effects. In the 1400-1600 nm region, the strong absorption by water and sugars leads to... with me s A sharp drop in correlation with physicochemical properties. During microwave drying, with me s The correlation between physicochemical properties and hot air drying across the entire wavelength range is similar, but the correlation curve is smoother, possibly due to the difference in dehydration mechanisms between the two drying methods. Previous studies have shown that chemical composition is related to the absorption of chemical groups within tissues, while scattering is related to the physical structure of the tissue. In this invention, the physicochemical content under the two drying methods and... with me s There is a relatively obvious negative correlation between them. This is because the decrease in moisture content and changes in chemical substances (such as starch gelatinization) during the drying process will cause changes in the tissue structure of the sample, thus leading to... with me s A linear increase in the relationship between physical and chemical properties.

[0082] like Figure 14 As shown, under the two drying methods, F em A relatively obvious negative correlation was observed with the three physicochemical indicators. Because fluorescence signals in complex matrices are affected by both absorption and scattering, the characteristic absorption of anthocyanins in the 500-600 nm range causes fluctuations in the correlation curve within this wavelength band. m a The even stronger "Black King Kong" F em The correlation with physicochemical properties is weaker, indicating that the strong-absorbing matrix has a stronger interference effect on fluorescence; at the same time, the reduction of water content is accompanied by changes in physical structure, thus water content also has a certain correlation with fluorescence signal across the entire wavelength range. with me s Consistent with the correlation analysis of physicochemical indicators, due to the difference in dehydration mechanisms between the two drying methods, the microwave drying process... F em The correlation with physicochemical indicators is less affected by absorption, and its correlation curve is smoother.

[0083] (5) Establishment and validation of quantitative prediction models for anthocyanins and total polyphenols The correlation analysis described above explains the relationships between various regular changes during the drying process. It should be noted that moisture content and colorimetric indicators are not involved in model establishment, but only in the correlation analysis. During the drying process, changes in moisture and colorimetry can be reflected in the absorption signal, while microstructure can be reflected in the scattering signal. Anthocyanins and polyphenols, on the other hand, exhibit fluorescence signals. The correlation analysis results, aided by moisture content and colorimetric indicators, verify the correlation between optical parameters and anthocyanins and total polyphenols during the drying process. Integrating absorption and scattering into the fluorescence signal can improve the prediction accuracy.

[0084] This invention selects wavelengths within the 380-780 nm band based on correlation analysis results. F em , m a ( m a-VIS ), within the 1100-1670 nm band m a ( m a-NIR ), in the 1100-1400 nm band with me s ( with me s-NIR As a characteristic optical band, it is used in Savitzky-Golay smoothing (SGS) for... m a , with me s and F em After the spectrum is smoothed, principal component analysis (PCA) is used to analyze each spectrum of the characteristic optical bands and extract the top 7 principal components (PCs) of each spectrum in different bands.

[0085] A quantitative prediction model for anthocyanins and total polyphenols was established based on data fusion at the data layer. This model is based on spectral compression without PCA. F em , m a-VIS , m a-NIR , with me s-NIR Based on the normalization of different spectra, a prediction model for anthocyanin and total polyphenol content was established using a Long Short-Term Memory (LSTM) network. The LSTM parameters were set as follows: 100 training cycles, 6 iterations per cycle, 64 and 32 hidden neurons, batch size of 32, and initial learning rate of 0.001.

[0086] A quantitative prediction model for anthocyanins and total polyphenols was established based on feature layer data fusion. Fem , m a-VIS , m a-NIR , with me s-NIR The first 7 PCs, and combinations of PCs after different normalization treatments, were used to establish a prediction model for anthocyanin and total polyphenol content using LSTM.

[0087] The two fusion methods mentioned above are denoted as LF and MF, which represent full-band fusion and fusion of the first 7 PCs, respectively.

[0088] As can be seen from Tables 1 and 2, the overall prediction performance of the anthocyanin quantitative model based on a single optical feature is mediocre, with a prediction set determination coefficient (...). R 2 p A score below 0.9 or some models exhibiting overfitting. Compared to models based on... m a and with me s The constructed LSTM model is based on the 380-780 nm band. F em The modeling method exhibits a relative advantage. Specifically, the coefficient of determination (COP) of the training set for the anthocyanin prediction model constructed during the hot air drying process is significantly higher. R 2 c The value is 0.866. R 2 p = 0.820, the constructed anthocyanin prediction model in microwave drying R 2 c = 0.892, R 2 p = 0.823, representing an improvement of approximately 1-35% compared to models built using other single optical characteristics. This indicates that selecting fluorescence optical characteristics specific to active substances as modeling parameters can improve the predictive specificity and accuracy of the model. F em and m a / with me s After fusion, the performance of the prediction models established based on dual optical features was improved to varying degrees. For the hot air drying process, the analysis error of the anthocyanin LSTM model established based on dual optical features (…) RPD All of them can reach 2 or above, and R 2 c and R 2p All achieved a coefficient of determination above 0.8; however, the performance of the dual-feature anthocyanin prediction model constructed for the microwave drying process was slightly worse. Although the coefficient of determination for the training set reached above 0.9, the performance for the prediction set decreased significantly. R 2 c and R 2 p The difference is large, indicating a risk of overfitting in the model.

[0089] Table 1. Prediction results of anthocyanin content in colored potatoes during hot air drying based on different optical characteristics.

[0090] Fusion F em , m a and with me s The anthocyanin quantitative models constructed using three optical features showed significantly improved predictive performance across different drying methods. Among them, feature layer fusion exhibited the best predictive performance during hot air drying. R 2 c and R 2 p The values ​​were 0.976 and 0.908, respectively, representing improvements of 11.0% and 8.8% compared to single fluorescence modeling. Meanwhile, the root mean square error of the training set (...) RMSEC ) and the root mean square error of the prediction set ( RMSEP Compared to single fluorescence modeling, the levels decreased by 30.645 mg / 100 g and 13.282 mg / 100 g, and the model... RPD The value remained stable above 3. In contrast, the model's prediction performance during microwave drying was significantly improved in the data layer fusion mode. R 2 c Reaching 0.982, R 2 p The value reached 0.930, representing improvements of 9% and 10.7% respectively compared to single fluorescence modeling. RMSEC and RMSEP The levels decreased by 23.841 and 15.263 mg / 100 g, respectively, and the model... RPD It also remains stable above 3.

[0091] Table 2. Prediction results of anthocyanin content in colored potatoes during microwave drying based on different optical characteristics.

[0092] Tables 3 and 4 show that the LSTM model based on a single optical feature exhibits a certain predictive ability for the total polyphenol content of colored potatoes. Specifically, based on... m a-NIR It can achieve optimal prediction results in a single optical characteristic, and the prediction model in the hot air drying process. R 2 c and R 2 p The values ​​were 0.910 and 0.854 respectively, while in microwave drying... R 2 c and R 2 p The values ​​were 0.895 and 0.875, respectively. The performance of models built based on single fluorescence characteristics is limited. This is partly due to the complexity of the total polyphenol system, which includes components with significant differences in thermal stability, such as anthocyanins, chlorogenic acid, and flavonoids, resulting in dynamic changes in fluorescence intensity and peak position during heating. Furthermore, the characteristic fluorescence spectra of the Maillard reaction products (melanoidins) highly overlap with those of phenolic substances, reducing the detection sensitivity of polyphenol fluorescence signals. The total polyphenol content showed a significant correlation with drying time and moisture loss rate, explaining the underlying mechanism behind the superior predictive performance of polyphenols based on the 1100-1670 nm moisture characteristic absorption band.

[0093] based on F em , m a and with me s A data-layer fusion LSTM model based on three optical features can further improve the accuracy of total polyphenol quantitative prediction. In a hot air drying system, the fusion model... R 2 c and R 2 p Compared to the optimal single optical feature model, the improvements were 7.0% and 7.4%, respectively. RMSEC and RMSEP The concentrations decreased by 152.895 mg / 100g and 80.152 mg / 100g, respectively; in the microwave drying system, the fusion model was superior to the optimal single optical feature model. R 2 c and R 2 p They increased by 9.4% and 4.7% respectively. RMSEC and RMSEPThe concentrations decreased by 186.563 mg / 100 g and 31.796 mg / 100 g, respectively. This result indicates that the multi-source optical information fusion strategy can effectively overcome the limitations of single-feature detection, achieving high-precision prediction of total polyphenol content during the drying process. Furthermore, it can be seen that data-layer fusion achieves better prediction results compared to feature-layer fusion. This may be due to the loss of some effective data during feature extraction. Therefore, an appropriate data fusion strategy should be selected based on the actual sample to achieve optimal prediction performance.

[0094] Table 3. Prediction results of total polyphenol content in colored potatoes during hot air drying based on different optical properties.

[0095] Table 4. Prediction results of total polyphenol content in colored potatoes during microwave drying based on different optical properties.

[0096] This invention is based on m a , with me s and F em The fusion method enables more accurate prediction of anthocyanin and total polyphenol content under both drying methods, with determination coefficients exceeding 0.9. This indicates that multi-source optical feature fusion can effectively characterize the dynamic changes of complex sample matrices. F em (380-780 nm) Specific acquisition of characteristic fluorescence signals of anthocyanins and polyphenols; m a-VIS (380-780 nm) can effectively eliminate the attenuation of fluorescence intensity caused by the internal filtering effect; m a-NIR (1100-1670 nm) can reflect the dynamic changes of moisture, starch and other chemical substances during the sample drying process; with me s-NIR (1100-1400 nm) can effectively characterize the changes in the physical structure of the sample during the drying process. Therefore, the fusion of multiple optical features can effectively improve the stability and accuracy of the dynamic prediction model of anthocyanin and total polyphenol content during the drying process.

[0097] The above examples are merely specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments and many variations are possible. All variations that can be directly derived or conceived by those skilled in the art from the disclosure of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A rapid, non-invasive method for detecting active ingredients during the drying process of colored potatoes, characterized in that, include: (1) Obtain samples of colored potato chips that have been processed by different drying processes and are at different degrees of dryness; (2) The absorption coefficient of potato slices in the 400-1700 nm wavelength range was measured using a dual integrating sphere system. μ a With reduced scattering coefficient μ' s Simultaneously, a laser-induced fluorescence system was used to measure the fluorescence intensity spectrum of potato slices in the 400-1000 nm wavelength band using a 375 nm wavelength laser as the excitation light. F em To obtain multiple optical characteristics; (3) Physicochemical tests were conducted on potato slices to obtain physicochemical indicators for anthocyanins and total polyphenols; (4) Based on the correlation between the multi-dimensional optical characteristics obtained in step (2) and the anthocyanin content and total polyphenol content measured in step (3), determine the characteristic optical bands that are significantly related to the anthocyanin content and total polyphenol content. (5) Based on the optical data in the characteristic optical band determined in step (3), establish a quantitative prediction model for the content of anthocyanins and total polyphenols, which is used for rapid and non-invasive detection of the active ingredients during the drying process.

2. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1, characterized in that, The drying process described in step (1) includes hot air drying and microwave drying; The hot air drying was carried out at temperatures of 60℃, 70℃, 80℃, and 90℃, and samples were taken at 0, 30, 60, 90, 120, 150, 180, and 210 minutes of drying, respectively. The microwave drying was carried out at power levels of 400W, 480W, 560W, and 640W, and samples were taken at drying times corresponding to 0-26 minutes by adjusting the transmission frequency.

3. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1, characterized in that, Before drying, the potato samples were washed and peeled, and sliced ​​into thin slices with a thickness of 3.5 ± 0.2 mm using a slicer. The sliced ​​samples were blanched in boiling water for 2 minutes to inhibit polyphenol oxidase activity. The blanched samples were immediately placed in deionized water, protected from light and allowed to stand at room temperature. After removing excess water from the surface with a clean absorbent cloth, the samples were weighed using an electronic balance.

4. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1, characterized in that, Step (4) includes: Calculate the absorption coefficient and reduced scattering coefficient in the 400-1700 nm band, and the correlation coefficients between the spectral data at each wavelength point of the fluorescence intensity spectrum in the 400-1000 nm band and the anthocyanin content and total polyphenol content. The continuous band with the highest absolute value of the correlation coefficient was selected as the characteristic optical band that was significantly correlated with the anthocyanin content and total polyphenol content.

5. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1 or 4, characterized in that, The characteristic optical bands determined in step (4) include: fluorescence intensity spectrum and absorption coefficient in the 380-780 nm band, absorption coefficient in the 1100-1670 nm band, and reduced scattering coefficient in the 1100-1400 nm band.

6. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1, characterized in that, Step (3) also includes determining the moisture content and color of the potato slices. The moisture content, color, multivariate optical characteristics, anthocyanin content, and total polyphenol content are used in the correlation analysis.

7. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1, characterized in that, During the physicochemical testing, the anthocyanin content was determined using the pH differential method, and the total polyphenol content was determined using the Folin-Ciocalteu colorimetric method.

8. The rapid, non-invasive detection method for active ingredients in the drying process of colored potatoes according to claim 1, characterized in that, The process of establishing and validating the quantitative prediction model includes: 1) Use Savitzky-Golay smoothing pairs μ a , μ' s and F em The spectrum was smoothed, and principal component analysis was used to analyze the 380-780 nm wavelength range. F em and μ a-VIS 1100-1670 nm band μ a-NIR and the 1100-1400 nm band μ' s-NIR Analysis was performed, and the top 7 principal components of various spectral bands were extracted; among them, μ a-VIS Indicates the wavelength range of 380-780nm μ a , μ a-NIR Indicates the wavelength range of 1100-1670 nm μ a , μ' s-NIR Indicates the wavelength range of 1100-1400 nm μ' s ; 2) Based on spectral compression without PCA F em , μ a-VIS , μ a-NIR , μ' s-NIR After normalizing and fusing different spectra, LSTM networks were used to establish prediction models for anthocyanin and total polyphenol content, respectively. 3) Based on F em , μ a-VIS , μ a-NIR , μ' s-NIR The first 7 principal components were normalized and fused after different spectra were analyzed, and LSTM network was used to establish prediction models for anthocyanin and total polyphenol content, respectively. 4) Compare the prediction effects of the various models obtained in steps 2) and 3) to determine the optimal prediction model for anthocyanins and total polyphenols.

9. According to the rapid non-invasive detection method for active ingredients in the drying process of colored potatoes as described in claim 8, the optimal prediction model for anthocyanin and total polyphenol content is: F em , μ a-VIS , μ a-NIR ,and μ' s-NIR The LSTM model after data fusion, where F em , μ a-VIS , μ a-NIR ,and μ' s-NIR Principal component analysis was not performed.

Images