CN109959751A - Thin-layer quantitative image recognition detection method for water-soluble natural products - Google Patents

Abstract

The invention discloses a thin-layer quantitative image recognition and detection method for water-soluble natural products, comprising: (1) spotting a sample to be detected containing water-soluble natural products on a thin-layer chromatographic plate; (2) using a first developing agent Carry out thin-layer chromatography step-by-step development on the thin-layer chromatography plate after spotting with the second developing agent; the first developing agent is a developing system composed of ethyl acetate/ethanol/water/acetic acid, and the second developing agent is A developing system consisting of chloroform/ethyl acetate/formic acid; (3) developing the color of the developed thin-layer chromatographic plate by using a dye system; (4) collecting the image of the developed thin-layer chromatographic plate, according to the The relationship between the image and the content of water-soluble natural products yields the content of water-soluble natural products. The method of the invention is easy to operate and does not need special instruments. The method of the invention has good quantification and high accuracy. The method of the invention has a fast test speed and can measure multiple samples simultaneously.

Description

Thin-layer quantitative image recognition detection method for water-soluble natural products

technical field

The invention belongs to the field of natural products and food and medicine, and particularly relates to a thin-layer quantitative image identification and detection method for water-soluble natural products.

Background technique

Water-soluble natural products, including flavonoids, flavonoid glycosides, anthocyanins, and saponins, are important active ingredients in plants, and have antibacterial and fresh-keeping, antioxidant, anti-aging, antibacterial, antiviral, hypolipidemic, and anticancer effects. The water-soluble natural products of flavonoids, flavonoid glycosides, and anthocyanins have excellent antioxidant effects, which can effectively balance free radicals in the human body by scavenging reactive oxygen species or their reactions to generate stable compounds, and their metal ion chelation ability can terminate free radicals. base chain reaction. A large number of in vivo and in vitro experiments and epidemiological data show that the water-soluble natural products of saponins have biological and pharmacological effects such as anti-inflammatory, lipid-lowering, anti-diabetic and cardiovascular disease (CVD) prevention. High-tech fields such as food and fine chemicals have broad application prospects. Commonly used detection methods for water-soluble natural products include spectrophotometric method and liquid phase method, but they require special instruments and are relatively troublesome to operate.

Recently, computer image processing technology has developed rapidly, and quantitative analysis of images has become feasible. Image processing programs represented by ImageJ have been widely used in the field of life science research, such as quantitative analysis of electrophoretic bands, cell counting, and quantitative processing of tissue structures. ImageJ is a public domain Java image processing program developed by the NIH. It can be run as an online applet or downloadable application on any computer with a Java 1.4 or higher virtual machine. Downloadable distributions are available for Windows, Mac OS, Mac OS X and Linux. It can display, edit, analyze, process, save and print 8-bit, 16-bit and 32-bit images. It can read many image formats including TIFF, GIF, JPEG, BMP, DICOM, FITS and "raw". It supports "stacks", a series of images that share a single window. It is multi-threaded, so time-consuming operations such as image file reading can be performed in parallel with other operations. It can calculate area and pixel value statistics for user-defined selections. It can measure distances and angles. It can create density histograms and line contour plots. It supports standard image processing functions such as contrast manipulation, sharpening, smoothing, edge detection and median filtering. It can perform geometric transformations such as scaling, rotation and flipping. All analysis and processing functions are available at any magnification factor. The program supports any number of windows (images) simultaneously, limited only by available memory. Spatial calibration can be used to provide real-world dimensional measurements in millimeters. Density or grayscale calibration is also possible. ImageJ is designed with an open architecture and provides extensibility through Java plugins. Custom acquisition, analysis and processing plugins can be developed using ImageJ's built-in editor and Java compiler. User-written plugins can solve almost any image processing or analysis problem.

Thin-layer analysis of natural products is a commonly used high-efficiency technique, but traditional thin-layer quantitative analysis requires special instruments and is inconvenient to use. Therefore, based on the latest computer image quantitative analysis methods, the thin-layer quantitative image recognition and detection of flavonoid polyphenols can simplify the operation and effectively broaden the scope of use. At present, there have been some studies using image recognition for quantitative detection of active ingredients. For example, the patent document with application number 200510115774.X discloses a thin-layer chromatography quantitative analysis method based on image processing technology, which includes the steps: (1) photographing a thin-layer plate carrying a sample to form a digital image; (2) pre-processing the photographed image Processing: lens distortion correction and noise filtering; (3) use the principle of uniform distribution of spot samples and equal vertical distance to determine control points; (4) construct the image background according to the control point interpolation method; (5) remove the background from the lens distortion corrected image , normalize the brightness of the pixel points according to the background; (6) segment the thin-layer image; (7) integrate and quantitatively analyze the gray level of the segmented area. However, this method only describes the general principles of image analysis, and cannot be directly applied to the quantitative detection of specific natural products. The patent document with the application number 201510163742.0 discloses a method for measuring the content of Morinda officinalis Nessaccharide. The trapezoidal integration method is used to integrate the chromatographic peaks, and quantitative analysis is performed according to the peak area and the sample concentration. But the method is limited to the determination of one specific sugar. The patent document of application number 201810451846.5 discloses a method for identifying the authenticity of honey, using ImageJ software to analyze oligosaccharides in honey samples. But this method is limited to the analysis of one class of sugars. Therefore, the development of quantitative analysis methods for different types of thin-layer image recognition is of great significance for the rapid analysis and identification of natural products.

SUMMARY OF THE INVENTION

The invention provides a thin-layer quantitative image identification and detection method for water-soluble natural products, which is simple in process, does not require special instruments, and has wide application.

A water-soluble natural product thin-layer quantitative image recognition detection method, comprising:

(1) spotting a sample to be detected containing a water-soluble natural product on a thin-layer chromatographic plate;

(2) using the first developing agent and the second developing agent to carry out thin-layer chromatography step-by-step development on the thin-layer chromatography plate after the spotting; the first developing agent is a developing system composed of ethyl acetate/ethanol/water/acetic acid , the second developing agent is the developing system that chloroform/ethyl acetate/formic acid is formed;

(3) using the dye system to develop the color of the expanded TLC plate;

(4) Image collection of the color-developed thin-layer chromatography plate, and the content of water-soluble natural products is obtained according to the relationship between the image and the content of water-soluble natural products.

Specifically, a thin-layer quantitative image recognition and detection method for water-soluble natural products, based on computer image recognition for simultaneous quantitative analysis of different water-soluble natural products (flavonoids, flavonoids, anthocyanins, saponins) and their mixtures, including the following step:

(1) Utilize ethyl acetate/ethanol/water/acetic acid and chloroform/ethyl acetate/formic acid developing solvent system to carry out thin-layer chromatography step-by-step development of different large classes of natural products;

(2) Utilize a unified dye system to carry out TLC color development for different water-soluble natural products;

(3) Electronization and quantitative analysis of water-soluble natural product thin-layer chromatographic images to obtain the final water-soluble natural product thin-layer quantitative result.

In the present invention, before detection, the water-soluble natural product can be dissolved in a solvent, prepared into a water-soluble natural product sample solution, and then developed by thin-layer chromatography; the water-soluble natural product sample solution has a concentration of 0.01-10 mg/mL ; further preferably 0.5-5 mg/mL. When thin-layer chromatography is used, the amount of the sample to be spotted is 0.1-10 microliters; preferably, it is 0.1-1 microliter. The water-soluble natural product sample to be detected can be dissolved in a variety of solvents, generally a solvent with good solubility and volatile nature is used. Preferably, the solvent is ethanol, dichloromethane, acetone, methanol or ethyl acetate. One or more of; further preferably one or both of ethanol or ethyl acetate.

The water-soluble natural product samples are spotted in thin-layer chromatography and then developed step by step in the developing solvent system, firstly using ethyl acetate/ethanol/water/acetic acid developing agent plate to develop to the distance of R1, then air-drying, and then Re-expand to R2 length with chloroform/ethyl acetate/formic acid developing solvent.

In the present invention, the R1 and R2 are the distances expressed by taking the initial spotting center point as the starting point and the frontmost line after the developing agent is developed as the end point. Preferably, the R1 is 15-25 mm; the R2 is 35-45 mm. Of course, for a specific deployment system, the deployment effects of the two deployment systems can also be controlled by controlling the deployment time. Preferably, the R1 is 20 mm; the R2 is 40 mm.

The volume ratio of each solvent in the first developing agent is ethyl acetate/ethanol/water/acetic acid=5:2.5～3/0.1～0.2/0.1～0.15; further preferred volume ratio is: ethyl acetate/ethanol/ Water/acetic acid=5/2.85/0.15/0.12; the volume ratio of each solvent in the second developing agent is chloroform/ethyl acetate/formic acid=1/0.8～1.2/0.01～0.05; further preferred volume ratio It is chloroform/ethyl acetate/formic acid=1/1/0.03.

After the water-soluble natural product sample is developed on the thin layer chromatography, the color is developed with a p-methoxybenzaldehyde color developing agent, the color developing time is 0.1-10 minutes, and the color developing temperature is 150-250°C.

The composition of the dye system is that the volume ratio of p-methoxybenzaldehyde/acetic acid/ethanol/sulfuric acid is 9.1-9.5/3.5-4/300-350/12.5. As a further preference, the p-methoxybenzaldehyde developer is composed of p-methoxybenzaldehyde/acetic acid/ethanol/sulfuric acid (9.2/3.75/338/12.5, v/v/v/v).

In the present invention, the water-soluble natural product includes one or more of flavonoids, flavonoid glycosides, anthocyanins, and saponins; as a further preference, the water-soluble natural product includes catechin, quercetin, tea saponin , one or more of cyanidin glucoside, oleanolic acid, cyanidin chloride, and rutin.

After the water-soluble natural product sample is colored on thin-layer chromatography, a camera or a scanner is used for electronic image acquisition.

As shown in Figure 1, after the color development is completed, a camera or a scanner can be used for electronic image acquisition to complete the electronic image of the water-soluble natural product sample. After the image of the water-soluble natural product sample is electronically digitized, an appropriate image processing software, such as ImageJ, etc., is selected to perform image optimization processing. Since the quantitative analysis is based on grayscale values, the images are first converted into grayscale images with software.

The image optimization processing of the water-soluble natural product sample includes denoising and smoothing. A series of operations such as removing salt and pepper noise, correcting, sharpening edges, and smoothing the above grayscale images are performed to make the images more suitable for quantitative analysis.

After the image optimization processing of the water-soluble natural product sample, background removal is performed. Select an appropriate algorithm to remove the interfering background in the image to be analyzed.

After the background of the water-soluble natural product sample image is removed, integration and quantification are performed to obtain a quantitative result.

Taking ImageJ software for image processing as an example, the process is as follows:

After the water-soluble natural product sample was developed on thin-layer chromatography, an HP M1005 scanner was used for electronic image acquisition, and the parameters were set as: millions of color output, resolution 300dpi, brightness and contrast default values, Pictures are saved in TIFF format.

After the images of the water-soluble natural product samples were electronically digitized, ImageJ software was used for image optimization.

First of all, the quantitative analysis is based on the gray value, so start the ImageJ software, first select "Image"-"Type"-"8-bit", convert the image to grayscale, and then select "Image"-"Adjust"-"Brightness" /Contrast...", click "Auto" in the pop-up dialog box to adjust the contrast and brightness of the image to make the area to be analyzed more visible.

Secondly, perform image optimization processing such as noise reduction and smoothing on the grayscale image, first select "Process" - "Noise" - "Despecle", and replace each pixel in the image with its 3*3 neighbors through a median filter. The median value of the pixels in the domain to remove the salt and pepper noise in the image, and then select "Process" - "Noise" - "RemoveOutliers..." to replace the pixels whose deviation exceeds the set threshold with the median value of the surrounding pixels to correct the image. ; Then choose "Process" - "Filters" - "UnsharpMask..." to extract the blurred version to sharpen and enhance the edges of the image; finally choose "Process" - "Smooth" to smooth the entire image to make the integral baseline more stable.

Then, after image optimization of the water-soluble natural product samples, ImageJ software was used for background removal. Select "Process"-"SubtractBackground...", based on the "Rollingball" algorithm in ImageJ, give a suitable rolling ball radius in the pop-up dialog box, and check "Lightbackground" and "Slidingparaboloid", the continuous image in the image Background removal.

After the background of the water-soluble natural product sample image was removed, ImageJ software was used to perform integration and quantification to obtain quantitative results. Click "RectangularSelectionTool" to select the target area, click "Ctrl+1" to determine the band, click "Ctrl+3" to integrate, use "StraightLineSelectionTool" to connect the two ends of the peak to form a closed graph, use "wandtool" to determine the peak area, complete Quantitative analysis.

The present invention can use common thin-layer chromatography plates of various specifications, and preferably, the standard curve is the same as the thin-layer chromatography plate used for actual detection.

When the present invention performs quantitative analysis, a standard curve between the integral value (peak area for ImageJ) and the sample content can be pre-built, and then, after the integral value is obtained, the content of the sample can be directly obtained, that is, the quantification of the sample to be detected can be realized. change.

A specific method for making a standard curve is as follows:

(i) Take a standard sample with a set concentration (for example, it can be 1.0, 2.0, 3.0 mg/L, and multiple groups can be set in parallel) with a set volume (can be 0.5 μl), and spot it on a thin-layer chromatographic plate. ;

(ii) use the first developing agent and the second developing agent to carry out thin-layer chromatography step-by-step development on the thin-layer chromatography plate after spotting; the first developing agent is a developing system composed of ethyl acetate/ethanol/water/acetic acid , the second developing agent is the developing system that chloroform/ethyl acetate/formic acid is formed;

(iii) using a dye system to develop the color of the expanded TLC plate;

(iv) image acquisition of the color-developed TLC plate to obtain an integral value;

(v) According to the method of steps (i) to (iv), the integrated values corresponding to the multiple groups of standard samples are obtained, and a standard curve between the sample concentration and the integrated value is prepared.

For the development distance of the first developing agent and the second developing agent, please refer to the above description of the detection method. For example, first use the ethyl acetate/ethanol/water/acetic acid developing agent plate to expand the distance to R1, then dry it, and then use trichloride. The methane/ethyl acetate/formic acid developer was re-developed to the length of R2. The R1 is 15-25 mm; the R2 is 35-45 mm. Of course, for a specific deployment system, the deployment effects of the two deployment systems can also be controlled by controlling the deployment time. Preferably, the developing agent system used in the actual detection is the same as the developing system used for making the standard curve.

Preferably, when constructing the standard curve, the expansion distance is the same as the expansion distance in the actual detection process, and the sampling conditions, color development conditions, and image acquisition and processing conditions used are consistent with the detection process.

The invention can be used for the detection of a single component sample, and can also be used for the detection of a mixed component sample. When used for the detection of a mixed component sample, the image acquisition unit is used to realize the image acquisition of each component sample point respectively, The concentration values of the samples are then calculated separately. Of course, by using the present invention, we can also realize the detection of multiple samples at one time. During detection, multiple samples can be spotted on the same thin-layer chromatographic plate, and then image acquisition and calculation are performed respectively. As shown in Fig. 1, the method of the present invention is adopted for the detection of six samples whose concentrations increase in sequence, namely, the samples are respectively spotted on the spotting positions on the same thin-layer chromatographic plate, and the distance between them can be as long as the interference can be avoided. ; Then unfold, carry out color development, and then scan with a computer scanner, read the corresponding images of each sample, integrate, and finally obtain the concentration value of the corresponding sample. The detection result is equivalent to the separate detection, and there is almost no interference.

Compared with the existing detection method, the present invention has the following advantages:

(1) The thin-layer quantitative image recognition and detection method for water-soluble natural products of the present invention is easy to operate and does not require special instruments.

(2) The method of the present invention, using the same development system, can detect a variety of water-soluble natural products, with good quantification, simple operation and high accuracy.

(3) The method of the present invention has a fast test speed, and can measure multiple samples at the same time; it can also detect samples containing multiple components, and has strong practicability.

Description of drawings

Fig. 1 is a process diagram of the method of the present invention for the detection of six samples whose concentrations are successively increased.

Detailed ways

The present invention will be described in further detail below with reference to the examples, but the embodiments of the present invention are not limited thereto.

The color developing agent adopted in the embodiment is p-methoxybenzaldehyde dyeing agent, and its composition is p-methoxybenzaldehyde/acetic acid/ethanol/sulfuric acid (9.2/3.75/338/12.5, v/v/v/v) .

First, create a standard curve as follows:

(i) Take a standard sample with a set concentration and spot it on a thin-layer chromatography plate (using the same thin-layer chromatography plate in Example 1); L five groups of standard samples (the samples are the same as the samples to be detected in each example), the solvent is ethanol, and the spotting volume is 0.5 microliters;

(ii) after drying, use the first developing agent and the second developing agent to carry out thin-layer chromatography step-by-step development on the thin-layer chromatography plate after the spotting;

Develop a 20 mm distance on a developer plate with ethyl acetate/ethanol/water/acetic acid (5/2.85/0.15/0.12, v/v/v/v), then air dry and reapply with chloroform/ethyl acetate / Formic acid (1/1/0.03, v/v/v) spreader re-expanded to a length of 40 mm;

(iii) using p-methoxybenzaldehyde dye to develop the color of the TLC plate after development: heating on a heating table at 250°C for 15 seconds to develop the color;

(iv) Scanning with a computer scanner after color development of the thin layer, the resolution is set to 300 dpi. The scanned images were denoised and smoothed with ImageJ, then the background was removed, and finally quantitative integration was carried out to obtain the integration values corresponding to five groups of standard samples of 0.5, 1.0, 2.0, 3.0, and 4.0 mg/L respectively;

(v) Prepare a standard curve between sample concentration and integrated value.

For the detection accuracy, for each group of standard samples, parallel detection was performed three times, and the average of the integrated values was taken as its corresponding integrated value.

For samples of different components, the standard curve needs to be drawn according to the above method.

In the process of making the standard curve, and after the color of the sample in the embodiment, the subsequent image processing process is as follows:

An HP M1005 scanner was used to collect electronic images of the color-developed TLC plates. The parameters were set as: millions of colors output, resolution 300dpi, default values of brightness and contrast, and pictures were saved in TIFF format.

Image optimization was performed using ImageJ software. First of all, the quantitative analysis is based on the gray value, so start the ImageJ software, first select "Image"-"Type"-"8-bit", convert the image to grayscale, and then select "Image"-"Adjust"-"Brightness" /Contrast...", click "Auto" in the pop-up dialog box to adjust the contrast and brightness of the image to make the area to be analyzed more visible.

Secondly, perform noise reduction and smoothing on the grayscale image, first select "Process" - "Noise" - "Despecle", and replace each pixel in the image with its 3*3 neighborhood through a median filter. The median value of the pixels to remove the salt and pepper noise in the image, and then select "Process" - "Noise" - "RemoveOutliers..." to replace the pixels whose deviation exceeds the set threshold with the median value of the surrounding pixels to correct the image; then Then select "Process" - "Filters" - "Unsharp Mask..." to extract the blurred version to sharpen and enhance the edges of the image; finally select "Process" - "Smooth" to smooth the entire image to make the integral baseline more stable.

Then, ImageJ software was used for background removal. Select "Process"-"SubtractBackground...", based on the "Rollingball" algorithm in ImageJ, give a suitable rolling ball radius in the pop-up dialog box, and check "Lightbackground" and "Slidingparaboloid", the continuous image in the image Background removal.

After the image background was removed, ImageJ software was used for integration and quantification, and the quantitative results were obtained. Click "RectangularSelectionTool" to select the target area, click "Ctrl+1" to determine the band, click "Ctrl+3" to integrate, use "StraightLine SelectionTool" to connect the two ends of the peak to form a closed graph, and use "wandtool" to determine the peak area, Get the corresponding integral value.

Example 1

The (+)-catechin ethanol solution was spotted with a capillary for 0.5 microliters to thin layer chromatography. /v/v) The developer plate was developed to a distance of 20 mm, then air-dried and re-developed to 40 mm with chloroform/ethyl acetate/formic acid (1/1/0.03, v/v/v) developer length. Then take out the solvent and use p-methoxybenzaldehyde dye to dye it after evaporation of the solvent, and place it on a heating table at 250°C for 15 seconds for color development. After the thin layer was developed, it was scanned with a computer scanner, and the resolution was set to 300 dpi. The scanned images were denoised and smoothed with ImageJ, and then the background was removed. Finally, quantitative integration was performed, and the integrated value was compared with the standard value to obtain the measurement result. The actual value was 1.20 mg/mL of a (+)-catechin solution, the measured value was 1.22 mg/mL, with a 2% error.

The thin-layer chromatography plate used was purchased from the market, with specifications (silica gel material, coating thickness: 0.2-0.25 mm, silica gel powder particle size: 10-40 microns).

Example 2

The quercetin ethanol solution was spotted with a capillary at 0.5 microliters to thin layer chromatography. ) developer plate was developed to a distance of 20 mm, then air-dried and re-developed to a length of 40 mm with chloroform/ethyl acetate/formic acid (1/1/0.03, v/v/v) developer. Then take out the solvent and use p-methoxybenzaldehyde dye to dye it after evaporation of the solvent, and place it on a heating table at 250°C for 15 seconds for color development. After the thin layer was developed, it was scanned with a computer scanner, and the resolution was set to 300 dpi. The scanned images were denoised and smoothed with ImageJ, and then the background was removed. Finally, quantitative integration was performed, and the integrated value was compared with the standard value to obtain the measurement result. The actual value of quercetin in ethanol solution was 1.51 mg/mL, the measured value was 1.53 mg/mL, and the error was 1.5%.

Example 3

The tea saponin ethanol solution was spotted with a capillary at 0.5 microliters to thin layer chromatography. After the solvent was evaporated, ethyl acetate/ethanol/water/acetic acid (5/2.85/0.15/0.12, v/v/v/v) The developer plate was developed a distance of 20 mm, then air-dried and re-developed to a length of 40 mm with chloroform/ethyl acetate/formic acid (1/1/0.03, v/v/v) developer. Then take out the solvent and use p-methoxybenzaldehyde dye to dye it after evaporation of the solvent, and place it on a heating table at 250°C for 15 seconds for color development. After the thin layer was developed, it was scanned with a computer scanner, and the resolution was set to 300 dpi. The scanned images were denoised and smoothed with ImageJ, and then the background was removed. Finally, quantitative integration was performed, and the integrated value was compared with the standard value to obtain the measurement result. The actual value of tea saponin ethanol solution of 2.5mg/mL, the measured value of 2.62mg/mL, the error is 5%.

Example 4

The cyanidin glucoside ethanol solution was spotted with a capillary for 0.5 microliters to thin layer chromatography. v/v) The developer plate was developed to a distance of 20 mm, then air-dried and re-developed to 40 mm with chloroform/ethyl acetate/formic acid (1/1/0.03, v/v/v) developer length. Then take out the solvent and use p-methoxybenzaldehyde dye to dye it after evaporation of the solvent, and place it on a heating table at 250°C for 15 seconds for color development. After the thin layer was developed, it was scanned with a computer scanner, and the resolution was set to 300 dpi. The scanned images were denoised and smoothed with ImageJ, and then the background was removed. Finally, quantitative integration was performed, and the integrated value was compared with the standard value to obtain the measurement result. The actual value of cyanidin glucoside solution was 1.50 mg/mL, and the measured value was 1.53 mg/mL, with an error of 2%.

Example 5

Oleanolic acid, quercetin, (+)-catechin, cyanidin chloride, rutin ethanol solution (the actual contents are: oleanolic acid: 1.21 mg/mL; quercetin: 1.31 mg/mL; (+)-catechin: 1.17 mg/mL; cyanidin chloride: 2.54 mg/mL; rutin: 1.12 mg/mL), spotted by capillary 0.5 μL to TLC After the solvent is evaporated, use ethyl acetate/ethanol/water/acetic acid (5/2.85/0.15/0.12, v/v/v/v) to spread out a distance of 20 mm on the developer plate, then air dry, and then use three The methyl chloride/ethyl acetate/formic acid (1/1/0.03, v/v/v) developer was re-developed to a length of 40 mm. Then take out the solvent and use p-methoxybenzaldehyde dye to dye it after evaporation of the solvent, and place it on a heating table at 250°C for 15 seconds for color development. After the thin layer was developed, it was scanned with a computer scanner, and the resolution was set to 300 dpi. The scanned image was denoised and smoothed with ImageJ, then the background was removed, and finally quantitative integration was performed. The integrated value was compared with the standard value to obtain the measurement result, and the content value was oleanolic acid: 1.17 mg/mL; quercetin cyanidin: 1.27 mg/mL; (+)-catechin: 1.15 mg/mL; cyanidin chloride: 2.46 mg/mL; rutin: 1.09 mg/mL. The error is less than 3%.

Claims (10)

1. a water-soluble natural product thin-layer quantitative image recognition detection method, is characterized in that, comprises: (1) The samples to be detected containing water-soluble natural products are spotted on a thin-layer chromatographic plate; (2) using the first developing agent and the second developing agent to carry out thin-layer chromatography step-by-step development on the thin-layer chromatography plate after the spotting; the first developing agent is a developing system composed of ethyl acetate/ethanol/water/acetic acid , the second developing agent is the developing system that chloroform/ethyl acetate/formic acid is formed; (3) using the dye system to develop the color of the expanded TLC plate; (4) Image collection of the color-developed thin-layer chromatography plate, and the content of water-soluble natural products is obtained according to the relationship between the image and the content of water-soluble natural products. 2. water-soluble natural product thin-layer quantitative image recognition detection method according to claim 1, is characterized in that, the volume ratio of each solvent in described first developing agent is ethyl acetate/ethanol/water/acetic acid=5/ 2.5～3/0.1～0.2/0.1～0.15; the volume ratio of each solvent in the second developing agent is chloroform/ethyl acetate/formic acid=1/0.8～1.2/0.01～0.05. 3. water-soluble natural product thin-layer quantitative image recognition detection method according to claim 1, is characterized in that, in step (2), at first utilize the first developing agent to expand to the distance of R1, then air dry, then use the first developing agent. The second developing agent is re-expanded to a distance of R2, the R1 is 15-25 mm; the R2 is 35-45 mm. 4 . The method for quantitative image recognition and detection of water-soluble natural product thin layers according to claim 1 , wherein the dye system is a p-methoxybenzaldehyde developer. 5 . 5. The method for quantitative image recognition and detection of water-soluble natural product thin layers according to claim 4, wherein the composition volume of the dye system is: p-methoxybenzaldehyde/acetic acid/ethanol/sulfuric acid=9.1～ 9.5/3.5～4/300～350/12.5. 6. The thin-layer quantitative image recognition and detection method for water-soluble natural products according to claim 1, wherein the concentration of the water-soluble natural products in the sample to be detected is 0.01-10 mg/mL, and the spotting amount is 0.1-10 mg/mL. 10 µl. 7 . The method for quantitative image recognition and detection of thin layers of water-soluble natural products according to claim 1 , wherein a camera or a scanner is used for electronic image acquisition. 8 . 8. water-soluble natural product thin-layer quantitative image recognition detection method according to claim 1, is characterized in that, utilizes ImageJ to process the collected image, and finally carries out integral quantification, utilizes the relation between integral value and sample content , to obtain the concentration of the sample to be tested containing the water-soluble natural product. 9. The water-soluble natural product thin-layer quantitative image recognition detection method according to claim 1, wherein a standard curve between the sample content and the integral value is constructed in advance, and during detection, according to the obtained integral value and the described The standard curve is used to obtain the concentration of the test sample containing the water-soluble natural product. 10 . The thin-layer quantitative image recognition and detection method for water-soluble natural products according to claim 1 , wherein the water-soluble natural products are one or more of flavonoids, flavonoid glycosides, anthocyanins, and saponins. 11 .