Ripeness degree estimation method and ripeness degree change prediction method

Abstract

To provide a ripeness degree estimation method for estimating a ripeness degree of fruits more appropriately than conventional methods.SOLUTION: A fruit peel, a test specimen, is irradiated with light in visible and near-infrared regions ranging from 400 to 1700 nm. Then, reflection intensity corresponding to the wavelength of the light is measured, and the reflectance is calculated from the reflection intensity. The calculated reflectance value is second-order differentiated to calculate a second-order derivative value. A ripeness degree index of the fruit is then estimated from the calculated secondary differential value on the basis of a ripeness degree prediction model. The ripeness degree prediction model represents a relation between the second derivative value of the reflectance and the ripeness degree index obtained by a sensory test. The ripeness degree prediction model is generated by partial least squares (PLS) regression analysis, with the ripeness degree index defined as a dependent variable and the second derivative of the reflectance defined as an independent variable.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention particularly relates to a method for estimating the ripeness of fruit in a non-destructive manner, and a method for predicting changes in ripeness due to storage temperature. [Background technology]

[0002] Conventionally, techniques for non-destructively measuring the ripeness of fruit have been reported. For example, Patent Document 1 describes a technology in which light from a light source S is irradiated onto a fruit F to be measured (irradiated light), and the intensity of scattered light from the fruit F is measured by a photodetector D from a direction at an angle θ to the axis of the irradiated light; in order to reduce the influence of scattered (reflected) light from the surface of the fruit F, which does not contain scattered information from inside the fruit F, the irradiated light is irradiated almost perpendicularly to the surface of the fruit F (so that the axis of the irradiated light passes almost through the center of the fruit F); the measurement optical axis is shifted to a position where reflected light from the surface does not enter the photodetector D; and the focal position of the lens is placed inside the object to be measured, thereby detecting mainly scattered light from inside, thereby measuring the firmness of the fruit as the degree of ripeness (hereinafter referred to as "prior art").

[0003] In the prior art, the "ripeness" of a fruit to be predicted is calculated as the firmness of the fruit. In addition, those skilled in the art may also interpret the difference in the ripeness of fruit as viscosity and elasticity. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2003-35669 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the case of fruits such as avocados (Persea americana), the firmness of the fruit is not necessarily proportional to the ripeness that humans actually perceive. In other words, the ripeness measured by conventional techniques does not necessarily match human sensations and is not appropriate. Therefore, there is a need for a more accurate method for estimating fruit ripeness.

[0006] The present invention has been made in view of the above circumstances, and aims to solve the above-mentioned problems. [Means for solving the problem]

[0007] The ripeness estimation method of the present invention involves irradiating the skin of a fruit to be tested with visible light of 400 to 1700 nm and light in the near-infrared region, measuring the reflection intensity corresponding to the wavelength of the light, calculating a reflectance from the reflection intensity, calculating a second-order derivative value by second-order differentiation of the calculated reflectance value, and estimating a ripeness index of the fruit from the calculated second-order derivative value based on a ripeness prediction model, wherein the ripeness prediction model is a model that shows the relationship between the second-order derivative value of the reflectance and the ripeness index obtained by a sensory test. The ripeness estimation method of the present invention is characterized in that the ripeness prediction model is generated by partial least squares (PLS) regression analysis, with the ripeness index as the dependent variable and the second derivative value of the reflectance as the independent variable. The ripeness estimating method of the present invention is characterized in that the fruit is an avocado (Persea americana). The ripeness change prediction method of the present invention is characterized in that it estimates the ripeness index of the fruit using the ripeness estimation method, measures the storage temperature of the fruit, and predicts the increase in the ripeness index of the fruit based on the estimated ripeness index and the measured storage temperature based on a ripeness change prediction model, and the ripeness change prediction model is a model that shows the relationship between the storage temperature and the increase in the ripeness index. [Effects of the Invention]

[0008] According to the present invention, a ripeness estimation method is provided that can appropriately estimate the ripeness of a fruit by irradiating the surface of the fruit to be tested with visible light of 400 to 1700 nm and light in the near-infrared region, measuring the reflection intensity, calculating the reflectance from the reflection intensity, calculating the second-order derivative value by second-order differentiation, and estimating the fruit ripeness index based on a ripeness prediction model that shows the relationship between the second-order derivative value of the reflectance and the ripeness index obtained by sensory testing. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a graph showing a ripeness index based on a sensory test according to an example of the present invention. [Figure 2] 10 is a graph showing the relationship between hardness and a ripeness index according to an embodiment of the present invention. [Figure 3] 4 is a graph showing the reflectance of each avocado fruit in the visible light to near-infrared light range according to an example of the present invention. [Figure 4] 4 is a graph showing second derivative values ​​of reflectance of each avocado fruit in the near-infrared light range shown in FIG. 3. [Figure 5] 1 is a graph showing the relationship between predicted and actually measured values ​​of a ripeness index according to an embodiment of the present invention. [Figure 6] 1 is a graph showing the relationship between the storage temperature and the increase in the maturity index per day in an example of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0010] <Embodiment> It is difficult to determine the optimum time to eat fruits such as avocados (Persea americana), and a non-destructive method for estimating the appropriate ripeness was needed. Furthermore, a standard for how ripeness changes with storage temperature was also needed. Therefore, there was a technical need for a non-destructive and easy method for measuring avocado ripeness and predicting the rate of ripeness increase at different storage temperatures. For this reason, the inventors of the present invention came up with the idea of ​​utilizing visible light and infrared light reflectance, and after extensive experiments, established a non-destructive method for estimating the ripeness of avocados and a method for predicting changes in ripeness due to storage temperature, thereby completing the present invention.

[0011] Specifically, as shown in the Examples below, the inventors measured the ripeness of avocados through a sensory test of taste (taste) by several monitors, using a 10-point index. Then, they nondestructively measured the reflectance in the visible and near-infrared regions between 400 and 1700 nm. Then, to determine the relationship between the measured reflectance and the ripeness index, they created a ripeness prediction model using partial least squares regression (PLS) analysis. This made it possible to measure the ripeness of avocados from the visible and infrared reflectance.

[0012] Furthermore, a significant correlation was found between the daily increase in avocado ripeness calculated by the PLS regression model and the storage temperature of the avocados. This made it possible to create a ripeness change prediction model. In other words, by using the PLS regression model to non-destructively measure the ripeness of avocados and then storing them at a specified temperature, it became possible to predict the optimal time to eat them.

[0013] More specifically, the fruit ripeness estimation method of this embodiment involves irradiating the surface of the fruit being tested with visible light of 400 to 1700 nm and light in the near-infrared region, measuring the reflection intensity corresponding to the wavelength of the light, calculating the reflectance from the reflection intensity, calculating a second-order derivative value by second-order differentiation of the calculated reflectance value, and estimating a fruit ripeness index from the calculated second-order derivative value based on a ripeness prediction model, which is characterized by being a model that shows the relationship between the second-order derivative value of the reflectance and the ripeness index obtained by a sensory test.

[0014] Specifically, the fruit in this embodiment may be various fruits such as avocado, acerola, apricot, strawberry, fig, plum, various citrus fruits, persimmon, quince, kiwi fruit, chestnut, guava, grapefruit, cherry, pomegranate, watermelon, star fruit, plum, Japanese pear, European pear, dragon fruit, durian, banana, pineapple, passion fruit, papaya, loquat, grapes, prunes, berries, mango, melon, peach, apple, etc. Of these, the fruits in this embodiment are particularly suitable for use with fruits whose taste (flavor) is not proportional to their hardness, viscosity, and elasticity, such as avocados, pears, mangoes, melons, peaches, and apples.

[0015] The maturity index according to this embodiment is characterized by being calculated by a sensory test of taste (gustatory sensation). Specifically, it is preferable to calculate the taste of fruits of the same variety by a "sensory test" in which the taste is evaluated by multiple people and an index is set from immature to fully ripe. If the number of people in this sensory test is about n=10 to 100, a ripeness prediction model can be suitably calculated.

[0016] In the method for estimating fruit ripeness according to this embodiment, it is possible to obtain continuous quantitative reflectance data by continuously irradiating the fruit with visible light in the range of 400 to 1700 nm and light in the near-infrared region at specific intervals of, for example, 1 to 100 nm using a commercially available small spectrophotometer or infrared spectrophotometer sensor used by those skilled in the art.

[0017] The ripeness prediction model according to this embodiment is characterized by being generated by PLS regression analysis in which the ripeness index is used as the dependent variable and the second derivative value of reflectance is used as the independent variable. Specifically, as described above, it is possible to generate a ripeness prediction model by performing a PLS regression analysis after second-order differentiation of the continuous reflectance data. In this case, in this embodiment, the ripeness index may be used as the dependent variable, and the second-order derivative of the reflectance may be used as the independent variable.

[0018] In addition, the ripeness change prediction method of this embodiment estimates the fruit ripeness index using the above-mentioned fruit ripeness estimation method, measures the storage temperature of the fruit, and predicts the increase in the fruit ripeness index based on the estimated ripeness index and the measured storage temperature based on the ripeness change prediction model, and is characterized in that the ripeness change prediction model is a model that shows the relationship between the storage temperature and the increase in the ripeness index. Specifically, the ripeness change prediction model according to this embodiment may be a model for the relationship between the storage temperature and the increase in the ripeness index per day. Also, this ripeness change prediction model may be expressed as a logarithmic approximation formula.

[0019] The above configuration can provide the following effects. Until now, it has been difficult to determine the optimum time to eat fruits such as avocados, and there has been a demand for a non-destructive technique for determining the appropriate ripeness. Conventional techniques have been used to create calibration curves for measuring ripeness non-destructively, using the fruit's hardness, viscosity, elasticity, etc. as indices of ripeness, but these do not necessarily match human senses and are therefore not appropriate.

[0020] In fact, as shown in Figure 2 in the Examples below, for example, in the case of avocados, the hardness and ripeness index were not proportional to each other as shown in the Examples. Specifically, in the case of avocados, although the ripeness index increases as the hardness decreases significantly, the ripeness index increases when the hardness is below 2 kgf / cm, which is the hardness at the general distribution stage. 2 The ripeness index was a mixture of optimal ripeness (5), immature (3), and overripe (7), and was not necessarily proportional to firmness. For this reason, it was thought that it would be difficult to judge the ripeness of fruits like avocados by firmness.

[0021] In contrast, the ripeness estimation method according to the present embodiment uses a ripeness prediction model that uses values ​​obtained from a sensory test as a ripeness index. Because it is humans who actually eat fruit, a more appropriate and accurate ripeness can be estimated by obtaining a ripeness index from a sensory test and using the obtained index to generate a model. In other words, by using a ripeness index determined by a sensory test that is perceived by humans, as in the present embodiment, a more appropriate and accurate fruit ripeness can be estimated non-destructively.

[0022] Furthermore, by using a ripeness prediction model generated from the second derivative of reflectance and a ripeness index from a sensory test, it is possible to easily and non-destructively determine ripeness. As shown in the examples described below, a significant correlation was observed at a level of 0.01% or higher between the predicted ripeness calculated using the ripeness prediction model generated by PLS regression analysis and the actual ripeness measured by a sensory test. In other words, a suitable correlation coefficient was obtained, allowing the generation of a practical model that can estimate with extremely high accuracy. Therefore, by applying the second derivative of reflectance described above to the calculated PLS regression ripeness prediction model, it is possible to measure the fruit ripeness index more appropriately, accurately, and non-destructively.

[0023] Furthermore, the ripeness estimation method according to the present embodiment can measure (calculate) reflectance using a commercially available small spectrophotometer, infrared spectrophotometer, or the like used by those skilled in the art. Therefore, the ripeness of fruit can be measured easily and non-destructively. In other words, no special equipment is required and inspection can be easily performed.

[0024] Furthermore, with conventional ripeness estimation methods, if the fruit was unripe at the time of measurement, it was necessary to store it for a specified period of time to allow it to ripen, and then non-destructively measure the ripeness as needed, before shipping and selling it at the appropriate time. This process was cumbersome and inaccessible to consumers who did not own measuring equipment. Therefore, there was a need for a technology that could accurately predict the relationship between storage temperature and the rate of ripeness at that temperature.

[0025] In contrast, the ripeness change prediction method according to this embodiment can predict the increase in the ripeness index of a fruit by measuring the estimated ripeness index and the storage temperature. That is, a method for predicting the increase in ripeness of a fruit such as an avocado is provided, and a standard for how ripeness changes depending on the storage temperature can be shown. This makes it possible to appropriately and accurately predict the change in ripeness depending on the storage temperature. Furthermore, it is possible to easily, appropriately, and accurately predict the increase in ripeness of a fruit (readiness). In fact, as shown in the examples described below, the ripeness change prediction model according to this embodiment also showed a significant correlation at a level of 0.01% or higher.

[0026] Furthermore, among fruits, avocados are highly nutritious and are being produced in increasing quantities worldwide. However, because avocados are easily damaged, timely and appropriately adjusted harvesting, post-harvest processing, and distribution are necessary. 99% of avocados distributed in Japan are imported, making quality control even more important. Therefore, the ripeness estimation method and ripeness change prediction method according to the present embodiment are expected to contribute to the distribution of high-quality avocados in the market.

[0027] Other Embodiments In the above embodiment, the ripeness of fruit is measured using a ripeness prediction model of PLS ​​regression analysis, but other statistical models and various machine learning methods such as neural networks and kernel machines can also be used. This neural network may use deep learning, etc.

[0028] Although the above-described embodiment uses the second derivative of the continuous reflectance, it is not necessary to use all wavelengths, and a specific discrete reflectance may be used. For example, as shown in the examples described below, reflectances at wavelengths at which peaks occur when the second derivative is taken, such as 560, 980, 1180, 1330, 1420, 1650, 1120, 1300, and 1380 nm, may be acquired and used to estimate a ripeness index.

[0029] In the above-described embodiment, an example has been described in which the ripeness change prediction model is used in the ripeness change prediction method. However, the ripeness change prediction model can also be used as an optimal storage temperature estimation method for estimating the optimal storage temperature according to the ripeness. That is, the temperature at which the product should be stored to achieve the optimal ripeness may be estimated based on the number of days until shipping, etc.

[0030] Furthermore, although the following examples use avocado as an example of fruit, the inventors have obtained preliminary results showing that similar models can be generated and similar predictions can be made for other fruits, such as mangoes and pears, whose hardness and ripeness are not necessarily related. Therefore, the fruit ripeness estimation method and ripeness change prediction method according to the present embodiment can be implemented in a variety of applications, not necessarily limited to avocados.

[0031] Next, the present invention will be further described by way of examples with reference to the drawings, but the present invention is not limited to the following specific examples. [Example]

[0032] Materials and Methods (Test material) The material used was commercially available lotus seeds of avocado (Persea americana).

[0033] (Method for measuring spectral reflectance) Visible light reflectance was measured using a compact spectrophotometer (Spectro1™, Variable, Inc., Chattanooga, TN, USA) at three randomly selected points on the avocado surface in the wavelength range of 400–700 nm at 10 nm intervals. Near-infrared reflectance was measured using a near-infrared spectrometer (NIR-S-G1, InnoSpectra Co. Hsinchu, Taiwan) at three randomly selected points on the avocado surface in the wavelength range of 900–1700 nm at 3.5 nm intervals.

[0034] (Method for measuring fruit firmness) After measuring the reflectance, the avocado was cut in half lengthwise, and the hardness of the edible part was measured using a fruit hardness tester (FHT-15, Guang Zhou Landtek Instruments Co., LTD, Guangzhou, China).

[0035] (Taste test of maturity index) The maturity index according to this embodiment will be described with reference to FIG. A taste test was conducted to determine the ripeness index for this example. Specifically, the fruits were judged on a 10-point scale, with "1" representing a hard, immature fruit, "5" representing an optimal fruit, and "10" representing an overripe fruit. The judgment was made by 5 to 10 people, and the average value was used as the ripeness index for this example.

[0036] (Generating a calibration curve showing the relationship between reflectance and ripeness index) Partial least squares (PLS) regression analysis was used to calculate the relationship between reflectance and ripeness index. In this example, the ripeness index was used as the dependent variable, and the second derivative of reflectance was used as the independent variable. A calibration curve for the PLS regression analysis was generated using the statistical software "Origin Pro 2024" (Lightstone Corp., Tokyo, Japan).

[0037] (Relationship between storage temperature and maturity index increase) The relationship between storage temperature and changes in the ripeness index in this example was investigated. After measuring the reflectance of the fruit using the above-mentioned method for measuring polarized reflectance, the fruit was stored at temperatures of 15, 20, 25, 30, and 35°C for two days, and the reflectance was measured each day. On each measurement day, the ripeness index was calculated using the relationship between the reflectance and the ripeness index described above. This was then summarized to calculate the relationship between storage temperature and changes in the ripeness index.

[0038] 〔result〕 FIG. 2 shows the relationship between firmness and the ripeness index according to this example. The horizontal axis of FIG. 2 is the firmness of the fruit (kgf / cm 2 ) and the vertical axis shows the ripeness index obtained by sensory testing.

[0039] The results showed that the relationship between firmness and the maturity index could be expressed as a power approximation, and a significant correlation was observed between the two. It was also revealed that as firmness decreased, the maturity index increased. On the other hand, the hardness is 2kgf / cm 2 In the surrounding area, the ripeness index in this example was sometimes about 5, which is the optimum ripeness, sometimes about 3, which is immature, and sometimes about 7, which is overripe. In other words, even though the hardness was the same, the ripeness in the sensory test was different. For this reason, it was thought that it would be difficult to judge ripeness by hardness.

[0040] Figure 3 shows the measurement results of spectral reflectance, and Figure 4 shows the second derivative of those values. Figures 3(a) and 4(a) show the reflectance of each fruit in the visible light range, and Figures 3(b) and 4(b) show the reflectance of each fruit in the near-infrared light range. The horizontal axis in both Figures 3 and 4 represents wavelength. The vertical axis in Figure 3 represents reflectance, and the vertical axis in Figure 4 represents the second derivative of reflectance.

[0041] According to the second derivative values ​​in Figure 4, a negative peak was observed around 560 nm in the visible light region, and positive peaks were observed around 980, 1180, 1330, 1420, and 1650 nm in the near-infrared region, and negative peaks were observed around 1120, 1300, and 1380 nm.

[0042] FIG. 5 shows the results of generating a prediction model (ripeness prediction model) by performing PLS regression analysis using the second derivative values ​​of the reflectance spectrum and the measured values ​​of the ripeness index in this example. The results of the analysis of the predictive performance of this maturity prediction model are shown in Table 1 below:

[0043] [Table 1]

[0044] As a result, the correlation coefficient (R 2) was 0.963, the slope was 1, the intercept (Bias) was almost 0, and the t-value was 3E+01. Thus, there was a very high and significant correlation between the predicted and measured values ​​of the maturity index in this example. In addition, the "ratio of the standard deviation of the predicted sample set to the standard error of the predicted value (RPD)," which indicates the analytical accuracy of the calibration curve (model), was 8.0. This indicates that the non-destructive prediction model for the maturity index in this example is a model that is "practical and has the potential to estimate with extremely high accuracy." These results show that measuring the reflectance of avocados in the visible and near-infrared light ranges makes it possible to determine their ripeness non-destructively.

[0045] FIG. 6 shows the results of generating a ripeness change prediction model for the relationship between storage temperature (preservation temperature) and the amount of increase in the ripeness index per day, using the model of FIG.

[0046] Thus, the relationship between storage temperature and the increase in the maturity index could be expressed by a logarithmic approximation, and there was a significant correlation between the two. Specifically, the higher the temperature, the greater the daily increase in the maturity index. When stored at temperatures below 12°C, the daily increase in the maturity index was zero. This revealed that immature fruits should be stored above 12°C, while fully ripe fruits should be stored below 12°C.

[0047] It goes without saying that the configurations and operations of the above-described embodiments are merely examples, and can be modified as appropriate within the scope of the present invention. [Industrial Applicability]

[0048] The present invention can be used industrially to estimate the ripeness of fruit and contribute to the distribution of high-quality fruit in the market.

Claims

1. Irradiating the epidermis of a test fruit with visible light of 400 to 1700 nm and light in the near-infrared region; measuring a reflection intensity corresponding to the wavelength of the light and calculating a reflectance from the reflection intensity; calculating a second-order derivative value by second-order differentiation of the calculated reflectance value; estimating a ripeness index of the fruit using the calculated second derivative value based on a ripeness prediction model; The ripeness prediction model is a model that shows the relationship between the second derivative value of reflectance and the ripeness index obtained by a sensory test. A method for estimating ripeness.

2. The maturity prediction model was generated by partial least squares (PLS) regression analysis using the maturity index as a dependent variable and the second derivative of the reflectance as an independent variable. The method for estimating ripeness according to claim 1 .

3. The fruit is an avocado (Persea americana). The method for estimating ripeness according to claim 1 .

4. The ripeness index of the fruit is estimated by the ripeness estimation method according to any one of claims 1 to 3, measuring the storage temperature of the fruit; predicting an increase in the ripeness index of the fruit based on the estimated ripeness index and the measured storage temperature based on a ripeness change prediction model; The maturity change prediction model is a model that shows the relationship between the storage temperature and the increase in the maturity index. A method for predicting changes in ripeness.

Images