Method for detecting moisture content of kidney beans based on hyperspectral reflectivity

By screening characteristic spectral bands using hyperspectral technology and establishing a random forest regression model, the problem of rapid and accurate detection of green bean moisture content was solved, enabling effective monitoring of the freshness and edible quality of green beans during storage and preservation.

CN118914102APending Publication Date: 2024-11-08JILIN ACAD OF VEGETABLE & FLOWER SCI
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410960138.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Existing technologies make it difficult to quickly and accurately detect the moisture content of green beans, which makes it difficult to maintain their freshness and edible quality during storage and preservation.

Method used

Using hyperspectral reflectance data, characteristic spectral bands were screened by principal component analysis, and a random forest regression model was combined to establish a model for detecting the moisture content of green beans.

Benefits of technology

It enables rapid and non-destructive detection of the moisture content of green beans, improving the accuracy of freshness monitoring and the maintenance of edible quality during storage and preservation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118914102A_ABST
    Figure CN118914102A_ABST
Patent Text Reader

Abstract

The water content is an important measurement index for estimating the freshness of kidney beans. The invention discloses a method for detecting the water content of the kidney beans based on spectral reflectivity. Comprising the following steps: (1) a data source is spectral reflectivity and actually measured kidney bean moisture content data, the spectral reflectivity is measured by using an ultraviolet-visible-near infrared spectrophotometer, and the determination of the kidney bean moisture content adopts a drying method II; and establishing a kidney bean moisture content detection model based on the kidney bean spectral reflectivity. And step 4, dividing the total number of samples into a training set and a test set according to a ratio of 2: 1, and calculating a determination coefficient R2 and a root mean square error RMSE so as to ensure the reliability of the established model. The method can quickly, effectively and nondestructively obtain the water content of the kidney beans, and has important significance for detecting the freshness of the kidney beans in the storage and sales processes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of agriculture, and in particular refers to a method for establishing a kidney bean moisture content detection model based on high spectral reflectance.

[0002] Background Field

[0003] Kidney beans are one of the most popular vegetables because of their fresh, thick pods, tender green color, and rich nutrition. However, fresh beans have tender tissues and high respiration intensity. After harvest, they lose water, causing the beans to swell, the pods to turn yellow and soften, and are prone to quality deterioration such as rust spots and wood fiber. In particular, wood fiber causes the bean tendons to grow and the tissues to become rough, which seriously affects their edible quality. Therefore, real-time detection of the freshness of kidney beans and maintaining the sensory quality and edible quality of fresh beans are the primary tasks of fresh bean storage and preservation.

[0004] Moisture content is an important indicator for judging the freshness of kidney beans. For kidney beans, appropriate moisture content is a key factor in maintaining their freshness, taste and nutritional value. Therefore, timely and accurate acquisition of moisture content is of great significance for the storage and preservation of kidney beans. Hyperspectral technology has been widely used in the rapid detection of crop physiological indicators due to its advantages such as high efficiency and non-destructiveness.

[0005] The present invention adopts principal component analysis to preprocess hyperspectral data, selects random forest regression model, and establishes a method for detecting moisture content of kidney beans based on hyperspectral reflectance. Summary of the invention

[0006] The present invention proposes a method for detecting the moisture content of kidney beans by constructing a model based on hyperspectral reflectance. The method can quickly and effectively detect the moisture content of kidney beans and provide key data for monitoring the freshness of kidney beans during the preservation process.

[0007] The present invention implements the invention by adopting the following technical solutions:

[0008] A method for detecting the moisture content of kidney beans based on the establishment of a model based on hyperspectral reflectance, using hyperspectral reflectance as the data source, screening the characteristic spectrum through principal component analysis to extract effective information bands, using reflectance data as variable input, taking the measured moisture content as the target, and using the random forest algorithm to establish a kidney bean moisture content detection model. The specific steps are:

[0009] Step 1, the data sources used in this method are hyperspectral data and measured crop moisture content data, the hyperspectral data is measured using an ultraviolet-visible-near infrared spectrophotometer, and the measured moisture content of kidney beans is measured using a drying method;

[0010] Step 2: The hyperspectral data is processed by principal component analysis to reduce the dimensionality. Principal component analysis can extract the spectral bands with concentrated features, thereby reducing the dimensionality to facilitate the subsequent model establishment.

[0011] Step 3, using the random forest model, a spectrum inversion model for estimating the moisture content of kidney beans was established;

[0012] Step 4: Divide the samples into training set and test set in a ratio of 2:1 and calculate the correlation coefficient R 2 and root mean square error RMSE to ensure the fairness and reliability of model training and evaluation

[0013] In step 1, when measuring the spectrum of kidney beans, its wavelength range is between 200-2500nm, the sampling interval is 1nm, and a white board is used for calibration before each spectrum collection. The spectrum values ​​are collected multiple times. After removing the abnormal values, a total of 150 samples are collected as the final spectrum of the test sample; in the actual measurement of the moisture content of kidney beans, the weight of kidney beans is first weighed as the fresh weight, dried at 105°C to constant weight, and the weight of kidney beans is weighed as the dry weight. Finally, the moisture content is calculated using the following formula:

[0014]

[0015] Where ω is the moisture content of kidney beans, m is the fresh weight of kidney beans, and m s is the dry weight of Phaseolus vulgaris.

[0016] Step 9: Calculate the coefficient of determination R 2 and RMS error

[0017]

[0018] Among them, yi is the reference value of the sample, is the predicted value of the sample, is the average value of the samples, and n is the number of samples. The beneficial effects of the present invention are:

[0019] By adopting the above technical solution, the moisture content of kidney beans can be detected efficiently and non-destructively by first using principal component analysis to extract the spectral calculation in the feature set and then using random forest to build a model. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is the original spectrum of the method for inverting the moisture content of kidney beans based on the hyperspectral technology of the present invention;

[0021] Figure 2 It is a simplified workflow diagram of the method for inverting the moisture content of kidney beans based on hyperspectral technology of the present invention;

[0022] Figure 3 This is a prediction model result diagram of the method for inverting the moisture content of kidney beans based on the hyperspectral technology of the present invention;

[0023] Figure 4 The invention relates to a spectral data acquisition system for a method for inverting the moisture content of kidney beans based on hyperspectral technology. DETAILED DESCRIPTION

[0024] The present invention is further explained by specific examples below. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0025] like Figure 1 Spectral analysis technology plays an important role in precision agriculture and plant physiology research. By analyzing the spectral data of common bean (in the wavelength range of 200nm-2500nm, the present invention aims to explore the correlation between spectral features and plant water content, in order to improve the accuracy and efficiency of agricultural water management. Common bean samples were grown under controlled conditions and spectral data were collected at different stages of the growth cycle. The spectrometer used collected data points every 1 nanometer, covering the wavelength range of 200 to 2500nm.

[0026] 1. Visible light region (400-700 nm): The spectrum in this region reflects the absorption peaks of chlorophyll, which mainly appear at 440 nm and 675 nm and are related to the health status and physiological activities of plants.

[0027] 2. Near-infrared region (700-1300nm): The reflectivity in this region is relatively high, which shows the scattering effect of water in the plant body on light, especially the obvious water absorption characteristics at 970nm, which is closely related to the water content in the intercellular space.

[0028] 3. Short-wave infrared region (1300-2500nm): Significant water absorption peaks are observed at 1450nm and 1930nm, which are particularly important for analyzing the water stress response of plants.

[0029] The workflow of this method is as follows Figure 2 As shown in the figure, the data sources used in this method are hyperspectral data and measured crop moisture content data. The hyperspectral data are obtained by ultraviolet-visible-near infrared spectrophotometer, and the moisture content of kidney beans is determined by drying method. The hyperspectral data are processed by principal component analysis to reduce the dimension. Principal component analysis can extract spectral bands in feature sets, thereby reducing the dimension to facilitate subsequent model establishment. A spectral inversion model for estimating kidney bean moisture content is established using a random forest model. The samples are divided into training set and test set in a ratio of 2:1, and the determination coefficient R is calculated. 2 And root mean square error RM SE to ensure the fairness and reliability of model training and evaluation.

[0030] The prediction model results of this method are shown in the figure Figure 3As shown in the figure. The model inverts the moisture content of kidney beans through hyperspectral data, which not only shows high-precision prediction results, but also its statistical parameters reflect the high reliability of the model. By applying hyperspectral technology, this method can achieve rapid and accurate estimation of kidney bean moisture content, especially providing a scientific basis for real-time monitoring of kidney bean preservation process. Determination coefficient R 2 The 0.961 and RMSE 0.044 ensure that the model has high detection accuracy for real samples. In addition, the tightness of the confidence intervals of the slope and intercept of the model indicates the stability and repeatability of the prediction.

[0031] In summary, the present invention uses hyperspectral technology to establish a model, which can realize the rapid and non-destructive detection of moisture content in the kidney bean preservation process, which is of great significance to the kidney bean preservation process.

Claims

1. A method for inverting the moisture content of kidney beans based on hyperspectral technology, characterized in that: include: Step 1, the data source used in this method is the high spectral reflectance and moisture content data of kidney beans, the spectral reflectance data is measured by ultraviolet-visible-near infrared spectrophotometer, and the moisture content of kidney beans is measured by drying method; Step 2: Spectral reflectance is processed by principal component analysis to reduce dimension. Principal component analysis can extract spectral bands in feature sets, reduce noise and redundant information in the data set, and effectively identify key spectral bands that affect moisture determination. Step 3, using the random forest model, a spectrum inversion model for estimating the moisture content of kidney beans was established; Step 4: Divide the total number of samples into training set and test set in a ratio of 2:1 and calculate the coefficient of determination R 2 And root mean square error RMSE to ensure the accuracy and reliability of the established model.

2. The method for inverting the moisture content of kidney beans based on hyperspectral technology according to claim 1 is characterized in that: The step 1 comprises: Step 5, when measuring the spectrum of kidney bean, the wavelength range is between 200-2500nm and the sampling interval is 1nm; Step 6: Before collecting the spectrum each time, calibrate with a whiteboard, collect the spectrum values ​​multiple times, remove the abnormal values, and collect a total of 150 samples as the final spectrum of the test sample; Step 7, when experimentally determining the moisture content of kidney beans, first weigh the kidney beans as fresh weight, dry them at 105°C to constant weight, weigh the kidney beans as dry weight, and finally calculate the moisture content using the following formula: Where ω is the moisture content of kidney beans, m is the fresh weight of kidney beans, and m s is the dry weight of Phaseolus vulgaris.

3. The method for estimating the moisture content of kidney beans based on hyperspectral technology inversion according to claim 1 is characterized in that: The step 2 comprises: Step 8: Use PCA principal component analysis to screen the spectrum preprocessing and extract the spectral characteristic wavelengths.

4. The method for estimating moisture content of kidney beans based on hyperspectral inversion according to claim 1, characterized in that: The steps 3 and 4 include: Step 9: Calculate the coefficient of determination R 2 And root mean square error RMSE Among them, yi is the reference value of the sample, is the predicted value of the sample, is the sample mean, and n is the number of samples.

Citation Information

Cited By

  • Estimation method for inverting onion leaf net photosynthetic rate based on hyperspectral reflectivity

    CN119290820A

  • Estimation method of net photosynthetic rate of onion leaves based on hyperspectral reflectance inversion

    CN119290820B