JPH11304698A - Method and device for measuring amylose content in rice - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable the highly accurate measurement of amylose content in polished rice by performing partial least square regression analysis on the basis of an actual content value obtained in the process of the chemical measurement of a reference value of amylose content and absorbance measured in the process of the near-infrared spectrometry of an amylose content value and obtaining a specific equation predicting amylose content, and performing computations. SOLUTION: Milled-rice samples are collected, and a value of amylose content to be a reference value in the case of performing regression analysis, etc., is measured from the collected polished-rice samples by a chemical analysis method. In parallel with the process of the chemical measurement of a reference value of amylose content, the polished rice is irradiated with near-infrared light to obtain the spectrum of the transmitted light and to measure absorbance at each wavelength. Next, letting the actual value of amylose content obtained by the chemical analysis method be an object variable and the absorbance by near-infrared spectrometry be a predictor variable, regression analysis is performed to obtain a conversion equation (calibration curve) with high correlation and few errors. Then the absorbance data on the rice samples by near-infrared spectrometry is applied to the calibration curve to evaluate the predictability of amylose content.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for measuring amylose content of rice and an apparatus for measuring amylose content of rice.
In particular, the present invention relates to a method for measuring amylose content of rice by near infrared spectroscopy and an apparatus for measuring amylose content of rice.

[0002]

2. Description of the Related Art Conventionally, wavelengths of 800 to 25
Irradiation with light of 00 nanometers (nm: m ^-9 = 1/1000 μm) (hereinafter referred to as “near-infrared light”), the physicochemical properties of the object to be measured are determined based on the degree of reflected light and transmitted light. An analysis method called “near infrared spectroscopy” for nondestructively measuring has been developed and used. In this near-infrared spectroscopy, it is possible to grasp the components of the object to be measured only by analyzing transmitted light or the like obtained by irradiating near-infrared light without performing chemical component analysis. For this reason, near-infrared spectroscopy
Applied to a method for estimating the chemical components of polished rice and brown rice, quality evaluation of rice by near-infrared spectroscopy is spreading to various fields of the rice industry using rice as a raw material or using rice.

[0003] The chemical quality of rice, particularly the content of moisture, protein, amylose and the like of milled rice affects the taste and physical properties of cooked rice, and generally, the lower the proportion of amylose in rice, the lower the content of rice.
It is known that cooked rice is soft, sticky and tasteful.

[0004] For this reason, water, protein,
If the content of many components such as amylose can be determined by the near-infrared spectroscopy described above, it is very preferable to evaluate the taste and quality of rice.

[0005]

However, since the evaluation of the content of polished rice requires that the prediction error be small,
Currently, rice quality items that can be estimated by near infrared spectroscopy are
Only water content and protein content. However, it is not possible to properly predict the taste of rice only by the water content and protein content of the polished rice grains.

In particular, it is difficult to measure the amylose content of milled rice, which greatly affects the taste and physical properties of cooked rice, by near-infrared spectroscopy. The reasons are as follows.

[0007] 1) The range of quality fluctuation of rice for general rice used for modeling the estimation of amylose in polished rice was not sufficient.

2) The reproducibility of the amylose measurement method by chemical analysis was low, and the error fluctuation of the standard amylose value was large.

3) The level of modeling techniques for estimating amylose content from near-infrared transmitted light of polished rice was low.

[0010] From these facts, there is no established method of measuring the amylose content in polished rice by near-infrared spectroscopy, and it is satisfactory as a technique for predicting the taste and quality of rice using near-infrared spectroscopy. There was nothing to be expected.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a method and an apparatus for accurately measuring the amylose content in polished rice by near-infrared spectroscopy. Is to provide.

[0012]

In order to solve the above-mentioned problems, a method for measuring the amylose content of rice according to the present invention comprises irradiating near-infrared light having a wavelength of 804 to 1079 nanometers on a polished rice sample. The absorbance at which external light is absorbed by the polished rice sample at each wavelength is measured, and the wavelength is 808 ± 4.
The absorbance at nm is x1, and the wavelength is 816 ± 4 nm.
The absorbance at 825 ± 4 nm is x3, the absorbance at 833 ± 4 nm is x4, the absorbance at 841 ± 4 nm is x5, the absorbance at 850 ± 4 nm is x6, and the wavelength is 858 ± 4 nm. Absorbance at x7
And the absorbance at a wavelength of 866 ± 4 nm is x8, the absorbance at a wavelength of 875 ± 4 nm is x9,
The absorbance at a wavelength of 883 ± 4 nm is x10, the absorbance at a wavelength of 891 ± 4 nm is x11, and the wavelength is 9
The absorbance at 00 ± 4 nm is x12 and the wavelength is 908
The absorbance at ± 4 nm is x13 and the wavelength is 917 ± 4
The absorbance at nm is x14 and the wavelength is 925 ± 4 nm
The absorbance at wavelength 933 ± 4 nm is x16, the absorbance at 942 ± 4 nm is x17, the absorbance at 950 ± 4 nm is x18, the absorbance at 958 ± 4 nm is x19, and the wavelength is 967 ± 4 nm. Absorbance at x20
And the absorbance at a wavelength of 975 ± 4 nm is x21, the absorbance at a wavelength of 983 ± 4 nm is x22,
The absorbance at a wavelength of 992 ± 4 nm is x23, the absorbance at a wavelength of 1000 ± 4 nm is x24, and the wavelength is 10
The absorbance at 08 ± 4 nm is x25, and the wavelength is 1017.
The absorbance at ± 4 nm is x26, and the wavelength is 1025 ± 4.
The absorbance at nm is x27 and the wavelength is 1033 ± 4 nm.
, The absorbance at a wavelength of 1042 ± 4 nm is x29, the absorbance at a wavelength of 1050 ± 4 nm is x30, the absorbance at a wavelength of 1058 ± 4 nm is x31, the absorbance at a wavelength of 1066 ± 4 nm is x32, and the wavelength is 1075 ± 4 nm. Absorbance at x33
By performing partial least squares regression analysis using the absorbances x1 to x33 as explanatory variables and the amylose content y (%) in the polished rice sample as an objective variable, the coefficient b1 is obtained.
−222 ≦ b1 ≦ 1790, the coefficient b2 is −544 ≦ b2 ≦ 2140, and the coefficient b
3 is −5100 ≦ b3 ≦ 1520, coefficient b4 is −103000 ≦ b4 ≦ 1190, and coefficient b5 is −74700 ≦ b5 ≦ 3550.
0 and the coefficient b6 is -52700 ≦ b6 ≦ 53
700 and the coefficient b7 is −104000 ≦ b7 ≦
5550, and the coefficient b8 is -80600 ≦ b8 ≦
29100, and the coefficient b9 is -60700 ≦ b
9 ≦ 51200, and the coefficient b10 is −107000
≦ b10 ≦ 5750, and the coefficient b11 is −837
00 ≦ b11 ≦ 33400, and the coefficient b12 is −5.
0600 ≦ b12 ≦ 58200, and the coefficient b13 is −
100000 ≦ b13 ≦ 6350, coefficient b14 is −84200 ≦ b14 ≦ 24600, coefficient b
15 is −60400 ≦ b15 ≦ 50700, the coefficient b16 is −98800 ≦ b16 ≦ 7430, and the coefficient b17 is −78700 ≦ b17 ≦ 3360.
0 and the coefficient b18 is -63900 ≦ b18 ≦ 50
200 and the coefficient b19 is −95400 ≦ b19 ≦
6070 and the coefficient b20 is −78000 ≦ b20 ≦
25600, and the coefficient b21 is -52300 ≦ b
21 ≦ 56000, and the coefficient b22 is −103000
≤ b22 ≤ 2870 and the coefficient b23 is -783
00 ≦ b23 ≦ 32400, and coefficient b24 is −6
2200 ≦ b24 ≦ 52300, and the coefficient b25 is
−99500 ≦ b25 ≦ 3690, coefficient b26 is −77800 ≦ b26 ≦ 34600, coefficient b
27 is −59800 ≦ b27 ≦ 52300, coefficient b28 is −103000 ≦ b28 ≦ 4090, and coefficient b29 is −80900 ≦ b29 ≦ 3020.
0 and the coefficient b30 is −56000 ≦ b30 ≦ 54
300 and the coefficient b31 is −1690 ≦ b31 ≦
2980, and the coefficient b32 is -1710 ≦ b32 ≦
1540, and the coefficient b33 is −1980 ≦ b
33 ≦ 1400, constant term b. -201000
≤ b. Assuming ≦ 1270000, the amylose content y (%) in the above-mentioned polished rice sample is expressed by the following equation: × x10 + b11 × x11 + b12 × x12 + b13 × x13 + b14 × x14 + b15 × x15 + b16 × x16 + b17 × x17 + b18 × x18 + b19 × x19 + b20 × x20 + b21 × x21 + b22 × x22 + b23 × x23 + b24 × x26 + b25 × x26 + b25 × x26 + b25 × x26 + b25 × x27 x32 + b33 x x33 + b. It is characterized by being calculated by:

In the method for measuring the amylose content of rice described above, preferably, the coefficient b1 = b2 = b3 = b31 =
b32 = b33 = 0, and the coefficient b4 is -26133
≤ b4 ≤ 957, and the coefficient b5 is -64.
320 ≦ b5 ≦ 25608, and the coefficient b6 is
−959 ≦ b6 ≦ 53699, and the coefficient b7
Is set to −21172 ≦ b7 ≦ 5550, and the coefficient b8 is set to −71496 ≦ b8 ≦ 21526,
The coefficient b9 is set to −3954 ≦ b9 ≦ 51172.
And the coefficient b10 is -20859 ≦ b10 ≦ 5
748, and the coefficient b11 is -72463 ≦ b11 ≦
26610, and the coefficient b12 is 3 ≦
b12 ≦ 58193, and the coefficient b13 is −201
04 ≦ b13 ≦ 6350, and the coefficient b14 is −
73737 ≦ b14 ≦ 18320, the coefficient b15 is −4475 ≦ b15 ≦ 50700, the coefficient b16 is −18813 ≦ b16 ≦ 7428, the coefficient b17 is −67999 ≦ b17 ≦ 25117, and the coefficient b18 is −6790 ≦ b18 ≦ 502
00, and the coefficient b19 is set to −22797 ≦ b19 ≦
5056, and the coefficient b20 is -70343 ≦ b
20 ≦ 18703, and the coefficient b21 is 3
3 ≦ b21 ≦ 55996, and the coefficient b22 is −2.
3377 ≦ b22 ≦ 2866, and the coefficient b23 is
−68306 ≦ b23 ≦ 24348, and the coefficient b
24 is −5034 ≦ b24 ≦ 52266, the coefficient b25 is −22473 ≦ b25 ≦ 3690, and the coefficient b26 is −68729 ≦ b26 ≦ 262.
65 and the coefficient b27 is -3992 ≦ b27 ≦
52300, and the coefficient b28 is −22176 ≦ b
28 ≦ 4090, and the coefficient b29 is −7006
8 ≦ b29 ≦ 21690, and the coefficient b30 is
0 ≦ b30 ≦ 54300, and the constant term b. −201000 ≦ b. ≤346450.

In the above-mentioned method for measuring the amylose content of rice, preferably, the coefficient b1 = b2 = b3 =
b31 = b32 = b33 = 0, and the coefficient b4 is -27.
13 ≦ b4 ≦ −68, and the coefficient b5 is
501 ≦ b5 ≦ 5442, and the coefficient b6 is
−959 ≦ b6 ≦ 2667, and the coefficient b7 is
−4059 ≦ b7 ≦ 5550, and the coefficient b8 is
−2586 ≦ b8 ≦ −658, the coefficient b9 is −3416 ≦ b9 ≦ −159, and the coefficient b10
Is set to −765 ≦ b10 ≦ 5158, and the coefficient b
11 is −2894 ≦ b11 ≦ 3586, the coefficient b12 is 2199 ≦ b12 ≦ 4866, the coefficient b13 is −215 ≦ b13 ≦ 6350, the coefficient b14 is −4814 ≦ b14 ≦ 2213, and the coefficient b15 is − 4020 ≦ b15 ≦ −2287,
The coefficient b16 is set to 2439 ≦ b16 ≦ 7428, and the coefficient b17 is set to 259 ≦ b17 ≦ 2271.
And the coefficient b18 is −4561 ≦ b18 ≦ −141
5 and the coefficient b19 is −1884 ≦ b19 ≦ 39
74 and the coefficient b20 is −1722 ≦ b20 ≦ −
732, and the coefficient b21 is 1612 ≦ b21 ≦
4287, and the coefficient b22 is −2239 ≦ b22 ≦
2866, and the coefficient b23 is 665 ≦ b23
≦ 1840, and the coefficient b24 is −3436 ≦ b
24 ≦ −884, and the coefficient b25 is −680 ≦
b25 ≦ 3690, and the coefficient b26 is 463
≦ b26 ≦ 3438, and the coefficient b27 is −217
6 ≦ b27 ≦ −445, and the coefficient b28 is −5.
95 ≦ b28 ≦ 4090, and the coefficient b29 is −2.
121 ≦ b29 ≦ 768, and the coefficient b30 is
0 ≦ b30 ≦ 2004, and the constant term b. To-
26592 ≦ b. ≦ 69.

In the method for measuring the amylose content of rice described above, preferably, the coefficient b1 = b2 = b3 =
b31 = b32 = b33 = 0, and the coefficient b4 is -2713
≦ b4 ≦ −67, and the coefficient b5 is 666 ≦
b5 ≦ 2473, and the coefficient b6 is −237 ≦ b
6 ≦ 1324, and the coefficient b7 is −518 ≦ b7
≦ 835, and the coefficient b8 is −1340 ≦ b8 ≦
−658, and the coefficient b9 is −842 ≦ b9 ≦
−159, and the coefficient b10 is −14 ≦ b10 ≦
675 and the coefficient b11 is −1157 ≦ b11 ≦ 1
64 and the coefficient b12 is 2199 ≦ b12 ≦ 262
4 and the coefficient b13 is 750 ≦ b13 ≦ 2223
The coefficient b14 is -2890≤b14≤-2692, the coefficient b15 is -3390≤b15≤-2287, the coefficient b16 is 2439≤b16≤3340, and the coefficient b17 is 1640≤b17≤2088, The coefficient b18 is set to −2458 ≦ b18 ≦ −1415, the coefficient b19 is set to −1884 ≦ b19 ≦ −1406, the coefficient b20 is set to −1533 ≦ b20 ≦ −732, and the coefficient b21 is set to 3080 ≦ b21 ≦ 4287. The coefficient b22 is -1639 ≦ b22 ≦ 1120, the coefficient b23 is 665 ≦ b23 ≦ 1446, the coefficient b24 is −1120 ≦ b24 ≦ −884, and the coefficient b25 is −680 ≦ b25 ≦ −87, The coefficient b26 is set to 463 ≦ b26 ≦ 1279, the coefficient b27 is set to −1264 ≦ b27 ≦ −445, and the coefficient b28 is set to −508 ≦ b28 ≦ 38. And, the coefficient b29 and 37 ≦ b29 ≦ 386, the coefficient b30 and 311 ≦ b30 ≦ 938, the constant term b. −47 ≦ b. ≦ −17.

Further, the rice amylose content measuring apparatus according to the present invention comprises a light irradiating section for irradiating near-infrared light having a wavelength of 804 to 1079 nm to a polished rice sample, An absorbance measuring unit for measuring the absorbance absorbed by the polished rice sample, an absorbance at a wavelength of 808 ± 4 nm is x1, an absorbance at a wavelength of 816 ± 4 nm is x2, an absorbance at a wavelength of 825 ± 4 nm is x3, and a wavelength of 833 ± 4 nm The absorbance at a wavelength of 841 ± 4 nm is x5.
And the absorbance at a wavelength of 850 ± 4 nm is x6, the absorbance at a wavelength of 858 ± 4 nm is x7,
The absorbance at a wavelength of 866 ± 4 nm is x8, the absorbance at a wavelength of 875 ± 4 nm is x9,
The absorbance at 83 ± 4 nm is x10, and the wavelength is 891.
The absorbance at ± 4 nm is x11 and the wavelength is 900 ± 4
The absorbance at nm is x12 and the wavelength is 908 ± 4 nm
The absorbance at wavelength 917 ± 4 nm is x14, the absorbance at 925 ± 4 nm is x15, the absorbance at 933 ± 4 nm is x16, the absorbance at 942 ± 4 nm is x17, and the wavelength is 950 ± 4 nm. Absorbance at x18
And the absorbance at a wavelength of 958 ± 4 nm is x19, the absorbance at a wavelength of 967 ± 4 nm is x20,
The absorbance at a wavelength of 975 ± 4 nm is x21, and the absorbance at a wavelength of 983 ± 4 nm is x22.
The absorbance at 92 ± 4 nm is x23 and the wavelength is 1000
The absorbance at ± 4 nm is x24, and the wavelength is 1008 ± 4.
The absorbance at nm is x25, and the wavelength is 1017 ± 4 nm.
Is x26, the absorbance at wavelength 1025 ± 4 nm is x27, the absorbance at wavelength 1033 ± 4 nm is x28, the absorbance at wavelength 1042 ± 4 nm is x29, the absorbance at wavelength 1050 ± 4 nm is x30, and the wavelength is 1058 ± 4 nm. Absorbance at x31
And the absorbance at a wavelength of 1066 ± 4 nm is x32, the absorbance at a wavelength of 1075 ± 4 nm is x33,
By performing partial least squares regression analysis using the absorbances x1 to x33 as explanatory variables and the amylose content y (%) in the polished rice sample as an objective variable, the coefficient b1 is calculated as-
222 ≦ b1 ≦ 1790, and the coefficient b2 is
−544 ≦ b2 ≦ 2140, and the coefficient b3 is
-5100 ≦ b3 ≦ 1520, coefficient b4
Is set to −103000 ≦ b4 ≦ 1190, the coefficient b5 is set to −74700 ≦ b5 ≦ 35500,
The coefficient b6 is set to −52700 ≦ b6 ≦ 53700, and the coefficient b7 is set to −104000 ≦ b7 ≦ 555.
0 and the coefficient b8 is -80600 ≦ b8 ≦ 29
100 and the coefficient b9 is -60700 ≦ b9 ≦
51200, and the coefficient b10 is −107000 ≦ b10 ≦
5750, and the coefficient b11 is -83700 ≦ b
11 ≦ 33400, and the coefficient b12 is −50600
≤ b12 ≤ 58200 and the coefficient b13 is -1000
00 ≦ b13 ≦ 6350, and the coefficient b14 is −8
4200 ≦ b14 ≦ 24600 and the coefficient b15
-60400 ≦ b15 ≦ 50700, coefficient b16 is −98800 ≦ b16 ≦ 7430, coefficient b
17 is −78700 ≦ b17 ≦ 33600, coefficient b18 is −63900 ≦ b18 ≦ 50200, and coefficient b19 is −95400 ≦ b19 ≦ 607.
0 and the coefficient b20 is −78000 ≦ b20 ≦ 25
600 and the coefficient b21 is −52300 ≦ b21 ≦
56000, and the coefficient b22 is −103000 ≦ b22 ≦
2870, and the coefficient b23 is -78300 ≦ b
23 ≦ 32400, and the coefficient b24 is −62200
≤ b24 ≤ 52300 and the coefficient b25 is -995
00 ≦ b25 ≦ 3690, and the coefficient b26 is -7
7800 ≦ b26 ≦ 34600, and the coefficient b27 is
−59800 ≦ b27 ≦ 52300, coefficient b28 is −103000 ≦ b28 ≦ 4090, coefficient b
29 is −80900 ≦ b29 ≦ 30200, coefficient b30 is −56000 ≦ b30 ≦ 54300, and coefficient b31 is −1690 ≦ b31 ≦ 298
0 and the coefficient b32 is −1710 ≦ b32 ≦ 1
540 and the coefficient b33 is −1980 ≦ b33 ≦
1400, constant term b. −201000 ≦ b. ≤
Assuming that the amylose content y (%) in the above-mentioned polished rice sample is 1270000, the following formula is used. x10 + b11 × x11 + b12 × x12 + b13 × x13 + b14 × x14 + b15 × x15 + b16 × x16 + b17 × x17 + b18 × x18 + b19 × x19 + b20 × x20 + b21 × x21 + b22 × x22 + b23 × x23 + b24 × x32 + b25 × x26 + b25 × x26 + b25 × x + 26 × x33 + b. And a data analysis unit that calculates the following.

[0017] In the rice amylose content measuring device described above, preferably, the data analysis section includes the coefficient b1.
= B2 = b3 = b31 = b32 = b33 = 0, and the coefficient b
4 is −26133 ≦ b4 ≦ 957, the coefficient b5 is −64320 ≦ b5 ≦ 25608, and the coefficient b6 is −959 ≦ b6 ≦ 536.
99 and the coefficient b7 is −21172 ≦ b7 ≦
5550, and the coefficient b8 is -71496 ≦ b
8 ≦ 21526, and the coefficient b9 is −395.
4 ≦ b9 ≦ 51172, and the coefficient b10 is −2.
0859 ≦ b10 ≦ 5748, and the coefficient b11 is
−72463 ≦ b11 ≦ 26610, and the coefficient b
12 is set to 3 ≦ b12 ≦ 58193, the coefficient b13 is set to −20104 ≦ b13 ≦ 6350, and the coefficient b14 is set to −73737 ≦ b14 ≦ 183.
20 and the coefficient b15 is −4475 ≦ b15 ≦
50700, and the coefficient b16 is -18813≤b
16 ≦ 7428, and the coefficient b17 is −6799
1 ≦ b17 ≦ 25117, and the coefficient b18 is −
6790 ≦ b18 ≦ 50200, and the coefficient b19 is
−22797 ≦ b19 ≦ 5056, and the coefficient b
20 is −70343 ≦ b20 ≦ 18703, the coefficient b21 is 33 ≦ b21 ≦ 55996, and the coefficient b22 is −23377 ≦ b22 ≦ 28
66 and the coefficient b23 is −68306 ≦ b23 ≦
24348, and the coefficient b24 is −5034 ≦ b
24 ≦ 52266, and the coefficient b25 is −2247
3 ≦ b25 ≦ 3690, and the coefficient b26 is −6.
8729 ≦ b26 ≦ 26265, and the coefficient b27 is
−3992 ≦ b27 ≦ 52300, and the coefficient b
28 is −22176 ≦ b28 ≦ 4090, the coefficient b29 is −700068 ≦ b29 ≦ 21690, and the coefficient b30 is 0 ≦ b30 ≦ 543.
00 and the constant term b. −201000 ≦ b. ≦ 3
As 46450, the amylose content y in the polished rice sample is calculated.

In the above-mentioned rice amylose content measuring apparatus, preferably, the data analysis section sets the coefficients b1 = b2 = b3 = b31 = b32 = b33 = 0 and sets the coefficient b4 to -2713 ≦ b4 ≦ −68, the coefficient b5 is set to be 501 ≦ b5 ≦ 5442,
The coefficient b6 is -959≤b6≤2667, and the coefficient b7 is -4059≤b7≤5550.
And the coefficient b8 is −2586 ≦ b8 ≦ −65
8 and the coefficient b9 is −3416 ≦ b9 ≦ −1
59 and the coefficient b10 is −765 ≦ b10 ≦ 5
158 and the coefficient b11 is -2894 ≦ b11 ≦
3586, and the coefficient b12 is 2199 ≦ b12 ≦
4866, and the coefficient b13 is -215≤b13
≤ 6350, and the coefficient b14 is -4814 ≤ b
14 ≦ 2213, and the coefficient b15 is −4020 ≦
b15 ≦ −2287, and the coefficient b16 is 2439
≤ b16 ≤ 7428, and the coefficient b17 is 25
9 ≦ b17 ≦ 2271, and the coefficient b18 is −45.
61 ≦ b18 ≦ −1415, and the coefficient b19 is −1
884 ≦ b19 ≦ 3974, and the coefficient b20 is −
1722 ≦ b20 ≦ −732, and the coefficient b21 is
1612 ≦ b21 ≦ 4287, and the coefficient b22 is
−2239 ≦ b22 ≦ 2866, the coefficient b23 is 665 ≦ b23 ≦ 1840, and the coefficient b24
Is set to −3436 ≦ b24 ≦ −884, and the coefficient b
25 is −680 ≦ b25 ≦ 3690, the coefficient b26 is 463 ≦ b26 ≦ 3438, the coefficient b27 is −2176 ≦ b27 ≦ −445, the coefficient b28 is −595 ≦ b28 ≦ 4090, and the coefficient b29 is −2121 ≦ b29 ≦ 768,
The coefficient b30 is set to 0 ≦ b30 ≦ 2004, and the constant term b. -26592 ≦ b. ≤ 69
The amylose content y in the milled rice sample is calculated.

In the rice amylose content measuring apparatus described above, preferably, the data analysis section sets the coefficients b1 = b2 = b3 = b31 = b32 = b33 = 0 and sets the coefficient b4 to -2713 ≦ b4 ≦ −67, the coefficient b5 is 666 ≦ b5 ≦ 2473, the coefficient b6 is −237 ≦ b6 ≦ 1324, and the coefficient b is
7 is −518 ≦ b7 ≦ 835, and the coefficient b8
−1340 ≦ b8 ≦ −658, the coefficient b9 is −842 ≦ b9 ≦ −159, and the coefficient b10 is
−14 ≦ b10 ≦ 675, and the coefficient b11 is −1
157 ≦ b11 ≦ 164, and the coefficient b12 is 21
99 ≦ b12 ≦ 2624, and the coefficient b13 is 75
0 ≦ b13 ≦ 2223, and the coefficient b14 is −2890
≦ b14 ≦ −2692, and the coefficient b15 is −3390 ≦
b15 ≦ −2287, and the coefficient b16 is 2439 ≦ b
16 ≦ 3340, and the coefficient b17 is 1640 ≦ b17
≦ 2088, and the coefficient b18 is −2458 ≦ b18 ≦
−1415, and the coefficient b19 is −1884 ≦ b19 ≦ −
1406, and the coefficient b20 is −1533 ≦ b20 ≦ −
732, and the coefficient b21 is 3080 ≦ b21 ≦ 42
87 and the coefficient b22 is -1639 ≦ b22 ≦ 112
0 and the coefficient b23 is 665 ≦ b23 ≦ 1446
The coefficient b24 is −1120 ≦ b24 ≦ −884, the coefficient b25 is −680 ≦ b25 ≦ −87, the coefficient b26 is 463 ≦ b26 ≦ 1279, and the coefficient b27 is −1264 ≦ b27 ≦ −445. The coefficient b28 is -508≤b28≤38, the coefficient b29 is 37≤b29≤386, the coefficient b30 is 311≤b30≤938, and the constant term b. −47 ≦ b. Assuming ≦ −17, the amylose content y in the milled rice sample is calculated.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, a method for measuring amylose content of rice according to one embodiment of the present invention will be described in detail with reference to the drawings.

The method for measuring the amylose content of rice according to the present embodiment (hereinafter referred to as “the present method”) includes a sample collecting process, an amylose content reference value measuring process, an amylose content value near-infrared spectroscopic measuring process, and a calibration process. It consists of four steps: a line analysis step and an actual measurement step.

The outline of the present method is as follows. 1) Collect a white rice sample (sample collection process). 2) Next, an amylose content value, which is a reference value for performing regression analysis or the like in a subsequent calibration curve analysis process, is measured from the collected polished rice sample by a chemical analysis method (amylose content reference value chemical measurement process). ). 3) In parallel with the above-mentioned amylose content reference value chemical measurement process, a polished rice sample is irradiated with near-infrared light, the spectrum of the transmitted light is obtained, and the absorbance at each wavelength is measured (amylose content value near red). Outside spectrometry measurement process). 4) Next, regression analysis was performed using the actual amylose content value obtained by the chemical analysis method as the target variable, and the absorbance by near-infrared spectroscopy as the explanatory variable, and a conversion equation with high correlation and small error (calibration) Line) (data analysis process). 5) Thereafter, the absorbance data of the rice sample by near-infrared spectroscopy is applied to the above calibration curve to evaluate the predictability of the amylose content (test evaluation process).

Hereinafter, each of these steps will be described in detail.

(1) Sample Collection Process First, the present method has the following features in the process of collecting a polished rice sample. As sample rice, Indian type rice and Java type rice were excluded and limited to Japanese type rice. As sample rice, 110 samples of polished rice samples of Japanese rice for general use were collected from all over Japan from Hokkaido to Kyushu. In addition to the above 110 samples, for comparison, a variety of varieties from semi-mochi rice to high amylose rice were collected for comparison.

Since the milled rice sample was collected as described above, the method of the present invention has a sufficiently wide variation in the quality of the milled rice sample, and the method for measuring the amylose content that is most suitable for Japanese rice. It has the advantage that it can be.

FIG. 1 shows the distribution of the amylose content of the polished rice samples of these 110 samples of general rice. As shown in the figure, the amylose content is 13-21%, d. b. And the width of the mutation was 8%.

The distribution of the amylose content of Japanese milled rice is 15 to 23%, d. b. (Mutation width 8%), or 18 to 23%, d. b. (Mutation width 5%) has been reported. Therefore, the milled rice sample tested this time has a variation width of distribution of 8%, so it has a sufficient range of quality fluctuation, and it can be said that it almost covers Japanese type rice. .

(2) Amylose Content Reference Value Chemical Measurement Process Next, an accurate value of the amylose content of each rice sample was determined by the "iodine coloration method". The method has the following features in this process. 1) The milled rice yield of the sample rice was adjusted to 90 ± 1%, and the variation in the amount of amylose in the sample due to the milled rice was reduced. 2) The conditions from the addition of the iodine reagent to the starch gelatinization solution obtained from the rice sample to the colorimetric determination after the addition of the iodine reagent were unified to a temperature of 25 ° C. and a time of 20 minutes to reduce fluctuations in the measured values.

(3) Amylose Content Value Near-Infrared Spectroscopy Measurement Process In parallel with the above-mentioned amylose content reference value chemical measurement process, measurement was carried out by near-infrared spectroscopy. The procedure and contents of the method will be described below.

1) Near-infrared spectrometry in this method is as follows.
Using the "permeation method", the polished rice sample was filled in a container called a "cell", and the measurement was performed in two cases, that is, when the cell width was 18 mm and when the cell width was 35 mm.

2) The near-infrared light irradiating the polished rice sample has a wavelength in the range of 808 to 1075 nanometers (nm). Light was used. These 33 wavelengths are 808n
m, 816 nm, 825 nm, 833 nm, 841 n
m, 850 nm, 858 nm, 866 nm, 875 n
m, 883 nm, 891 nm, 900 nm, 908 n
m, 917 nm, 925 nm, 933 nm, 942 n
m, 950 nm, 958 nm, 967 nm, 975 n
m, 983 nm, 992 nm, 1000 nm, 1008
nm, 1017 nm, 1025 nm, 1033 nm, 1
042 nm, 1050 nm, 1058 nm, 1066 n
m, 1075 nm.

3) By measuring the light intensity, the intensity I of the near-infrared light of each of the above wavelengths before transmission through the sample. Then, the intensity It of the light after passing through the sample is obtained. The former I. The common logarithm log (I./It) of the ratio of the latter to It is referred to as absorbance. In general,
Since the absorbance is proportional to the concentration of the substance through which light passes and the optical path length, the absorbance in the present method is an index proportional to the amylose content (concentration of amylose in the sample:%). The wavelength change of the absorbance is called an absorption spectrum. FIG.
An example of the absorption spectrum of a polished rice sample according to the present method is shown below.

(4) Data analysis process The amylose content reference value obtained by the chemical analysis method in the above-mentioned amylose content reference value chemical measurement process is used as the target variable, and the amylose content value near-infrared spectrometry measurement process is performed. PLS regression analysis is performed using the absorbance x measured by infrared spectroscopy as an explanatory variable, and a prediction formula (regression formula or calibration curve) for predicting the amylose content from the near-infrared spectrum is obtained.

Conventionally, as a technique for this, a method called “Multiple Linear Regression Analysis (MLR)” has been used.
lysis)), but in this method, a more accurate PLS regression analysis method is used. Hereinafter, these will be described.

1) Linear multiple regression analysis method The objective variable is y, and the explanatory variables are x1, x2, x3,.
xd, the model equation of the multiple regression analysis is given by the following equation (1)

(Equation 1) It is represented as

In the above equation (1), b. Is a constant term, bk is a partial regression coefficient for xk, and e is a residual. The part excluding e from the above equation (1) is a regression equation (prediction equation or calibration curve). The regression analysis is an analysis method for obtaining the coefficient bk. First, for the objective variable y, the explanatory variable x
Is described (single regression model), and then a general multiple regression analysis method (multiple regression model) is described. After that, the PLS regression analysis method will be described.

A) Simple regression model The simple regression model formula is as follows: y = b. + X · b + e (2) The part excluding e from the above equation (1) is a regression equation (prediction equation or calibration curve).

When a set of x and y is obtained for n samples, the above equation (2) can be expressed as the following equation (3). y1 = b. + X1.b + e1y2 = b. + X2 .b + e2... Yn = b. + Xn b + en (3)

Here, the optimal section b. "Least Squares Method" to minimize the sum of squares of the residuals e1 to en of n samples in order to obtain
Is used. The sum of squares G of the residual is represented by the following equation (4).

(Equation 2)

G is b. Since G is a function of b, G is set to b in order to find the minimum value of G. When the values obtained by partial differentiation with respect to b and b are set to 0 and rearranged, the following simultaneous equations are obtained.

[0041]

(Equation 3)

[0042]

(Equation 4)

These simultaneous equations (5) and (6) are called normal equations.

Here, by using a vector and a matrix, it can be expressed as follows. Here, Y is a target variable vector, X is an explanatory variable matrix, and B is a coefficient vector.

[0045]

(Equation 5)

[0046]

(Equation 6)

[0047]

(Equation 7)

[0048]

(Equation 8)

Using these vectors and matrices, the normal equations (5) and (6) can be expressed as the following equation (11).

(Equation 9)

Here, it can be expressed as in the following equation (12).

(Equation 10) X ^T is a transposed matrix of the explanatory variable matrix X.

Therefore, the normal equations (5) and (6)
Can be expressed in the form of a matrix as X ^T XB = X ^T Y (13).

In the above equation (13), if X ^T X is a regular matrix, that is, if an inverse matrix of X ^T X exists, then (X ^T X) ⁻¹ X ^T XB = (X ^T X ^{) -1 X T Y ...... (14} ) B = (X T X) -1 X T Y ...... ( by multiplying both sides of the 15) (13) and (X ^T X) ^-1 from the left side of the formula (15) is obtained. The value of each element of the coefficient vector B can be obtained from the equation (15). Therefore, regression equation (prediction equation or calibration curve)
Can be obtained.

Here, (X) on the right side of the above equation (15)
^T X) ⁻¹ X ^T is the general inverse of matrix X.

2) Multiple regression model A model having two or more explanatory variables is called a multiple regression model, and the model formula of the above equation (1) is applied. The part excluding e from equation (1) is a regression equation (prediction equation or calibration curve). If a set of (x1, x2,..., Xd) and y is obtained for n samples, the model equation (1) becomes the following equation (1).
It is represented as 6). y1 = b. + X11 · b1 + x12 · b2 + x1d · bd + e1 y2 = b. + X21.b1 + x22.b2 + x2d.bd + e2... Yn = b. + Xn1 · b1 + xn2 · b2 + xnd · bd + en (16)

Here, the optimal section b. And slopes b1 and b2
, ..., bd, the residual e1 of n samples
The "least squares method" of minimizing the sum of squares of .about.en is used. The sum of squares G of the residual is represented by the following equation (17).

[Equation 11]

G is set to b in order to find the minimum value of G. , B
The partial differential of 1, b2,..., Bd is set to 0, rearranged, and a normal equation is obtained. Also, it is expressed as follows using a vector and a matrix. Where X is an explanatory variable matrix, Y
Is an objective variable vector, and B is a coefficient vector.

[0057]

(Equation 12)

[0058]

(Equation 13)

[0059]

[Equation 14]

Using these vectors and matrices and letting the transposed matrix of matrix X be X ^T , the normal equation can be expressed in the form of a matrix as X ^T XB = X ^T Y (21) .

Therefore, in the above equation (21), X ^T
X is regular matrix, i.e. if the inverse matrix of X ^T X is present, to obtain the value of the coefficient vector B by the following equation ^{(22) B = (X T} X) -1 X T Y ...... (22) The regression equation (prediction equation or calibration curve) can be obtained.

The multiple correlation coefficient R is used to test the significance of the regression equation (prediction equation or calibration curve) obtained by the least squares method. R is defined by the following equation (23).

(Equation 15)

Further, the square value R ² of R is often used as an explanation variance. The value of R ² takes a value from 0 to 1,
A value closer to 1 indicates that a better regression equation (prediction equation or calibration curve) was obtained.

The standard deviation s is also used as an index indicating the accuracy of the regression equation (prediction equation or calibration curve).
s is defined by the following equation (24).

(Equation 16)

However, the above-mentioned linear multiple regression analysis has the following problems. If there is a high correlation between the explanatory variables x (they have collinearity), the partial regression coefficient bi obtained by the multiple regression analysis becomes unstable and does not become an accurate prediction formula. When the number d of the explanatory variables x increases and becomes larger than the number n of samples, the partial regression coefficient bi cannot be uniquely obtained.

As new chemometrics for solving the problems of these multiple regression analysis methods, a "principal component regression analysis method" and a "PLS regression analysis method" described below are used.

2) Principal component regression analysis method This is a regression analysis method using a small number of uncorrelated principal components extracted from original variables as explanatory variables. This principal component regression analysis has the following advantages. Since the principal components are uncorrelated, the problem of "collinearity", which was a problem in multiple regression analysis, should not occur. Since the number of principal components is smaller than the number of original variables, the number of samples for preparing a calibration curve should be small.

However, since there is a problem of which principal component is selected as an explanatory variable, from a practical viewpoint, a PLS regression analysis method which will be described later is mostly used.

3) PLS Regression Analysis PLS (Partial Least Squares) regression analysis (partial least squares regression analysis) does not use the explanatory variable x as it is in the regression analysis, but uses a linear combination of the explanatory variable x. This is a method of performing a regression analysis using a variable t, t as an explanatory variable, q as a coefficient, and y = tq + e (25).

The PLS regression analysis has the following advantages. Since no inverse matrix is included, the problem of “collinearity” does not occur. Even if the number of explanatory variables exceeds the number of samples,
Models can be built. Since the information of the explanatory variable x is sequentially used through the latent variable t, the predictability can be examined while changing the degree of freedom of the PLS model. For this reason, the PLS regression analysis generally provides higher prediction accuracy than the principal component regression analysis, and is applied to the calibration of analytical chemistry.

In the following, when there is only one target variable y, a modeling method using the PLS regression analysis method (hereinafter referred to as “PLS
One analysis ". ) Will be described. When the relationship between the explanatory variable matrix X and one objective variable y is modeled by PLS regression analysis, the PLS model has the following 2
It consists of two basic expressions (26) and (27).

[0072]

[Equation 17]

[0073]

(Equation 18)

In the above equation (26), X is an explanatory variable matrix (element: x), A is the number of PLS components (the number of factors),
t _a is a th latent variable, p _a is the a-th loading, E is X the residual vector (element: e), T is latent variable vector (element: t), P is the loading vector (element: p) , ^PT are the transposed vectors of P. In the above equation (27), y is a target variable, q _a is a coefficient, Q is a coefficient vector (element: q), and e is a residual of y.

For simplicity, a one-component PLS model will be described first. In this case, equations (26) and (2)
7) can be expressed as in the following equations (28) and (29).

X = TP ^T + E (28)

Y = TQ + E (29)

In the multiple regression analysis method, the information of X is used as it is for modeling, but in the PLS regression analysis method, a part of the information of X is expressed through a latent variable vector T, which is used for modeling the target variable vector Y. Used for

In the above equation (29), Y is a simple regression analysis using the latent variable T.

The latent variable T is represented by a linear combination of the explanatory variable matrix X and the PLS weight vector W as shown in the following equation (30). T = XW (30)

Here, the PLS weight vector W is expressed by the equation (3)
As shown in 1), the norm is 1.

[Equation 19]

The criterion for constructing the PLS model is that the objective variable Y
And at the same time, increase the variance of T and use as much information as possible in X. For this purpose, the inner product S = Y ^T T, which is the covariance of Y and T, is considered, and the case where S is maximum is considered as the optimal model of PLS. Here, since S is maximized under the constraint that the norm of the PLS weight vector W is set to 1, the “Lagrange undetermined multiplier method” is used. Let μ be an undetermined constant, and consider a function G expressed by the following equation (32), and give a constraint condition that the norm of W is 1.

[0083]

(Equation 20)

Since the function G is a function of the variable W (element wk), G is partially differentiated with wk and set to 0. From the condition that the norm of the vector W = 1, the vector W is expressed by the following equation (33). Required by

(Equation 21)

The loading vector P is given by the following equation (3) from the condition that the sum of squares of the elements of the residual vector E of X is minimum.
4). ^{P = X T T / (T} T T) ...... (34)

The coefficient vector Q is given by the following equation (3) from the condition that the sum of squares of the elements of the residual vector E of Y is minimum.
5). ^{Q = Y T T / (T} T T) ...... (35)

In the PLS method, PL which is the number of latent variables
The model can be complicated by increasing the number of components of S. After the first PLS component is obtained as described above, the two-component PLS model formula is given by the following formulas (36) and (3)
7).

X = T ₁ P ₁ ^T + T ₂ P ₂ ^T + E (36)

Y = T ₁ Q ₁ + T ₂ Q ₂ + E (37)

In the PLS model with one component, T out of X
₁ P ₁ ^T is used to model Y, and T ₁ Q _{1 of} Y
Has been described, the remaining X and Y information can be expressed as in the following equations (38) and (39).

X _new = X−T ₁ P ₁ ^T (38)

Y _new = Y−T ₁ Q ₁ (39)

Assuming that the second latent variable is T ₂ , the following equation (40)
Consider a linear combination of X _new as follows. T ₂ = X _new · W ₂ = (X−T ₁ P ₁ ^T ) · W ₂ (40)

Here, W ₂ is a PLS weight vector. The criterion for PLS is the covariance of Y _new and T ₂ , Y ^T _new T
₂ is to determine W ₂ so as to maximize. From this, W ₂ is obtained by the following equation (41).

(Equation 22)

Hereinafter, similarly to the case of the first PLS component, T ₂ can be similarly calculated by the following equation (42) from the condition that minimizes the sum of squares of the elements of the residual vector E. T ₂ = X _new W ₂ = (X−T ₁ P ₁ ^T ) W ₂ (42)

Further, the second loading vector P ₂ is
It can be calculated by the following equation (43). P ₂ = X ^T _new T ₂ / (T ₂ ^T T ₂ ) = (X−T ₁ P ₁ ^T ) ^T T ₂ / (T ₂ ^T T ₂ ) (43)

The second coefficient vector Q ₂ is given by the following equation (4
4). Q ₂ = Y ^T _new T ₂ / (T ₂ ^T T ₂ ) = (Y−T ₁ P ₁ ^T ) ^T T ₂ / (T ₂ ^T T ₂ ) (44)

Hereinafter, similarly, the regression analysis of the PLS model when the number of components is 3 or more can be performed.

In the PLS regression analysis, the calculation method may be different depending on the data to be analyzed. These cases are as follows. When PLS analysis is performed on the objective variable Y and the explanatory variable X as they are. When auto-scaling the objective variable Y and the explanatory variable X. When the average value of the objective variable Y and the explanatory variable X is subtracted (subtracted) in advance.

Such a PLS regression analysis is performed by a computer. (Case in which the average value of Y and X is subtracted in advance) will be outlined below.

In this case, the above equation (26),
The basic expressions corresponding to (27) are the following expressions (45) and (46)
become that way.

[0102]

(Equation 23)

[0103]

(Equation 24)

Here, each column of the average value matrix of X is expressed by the following equation (47).

(Equation 25) It is the average value of each explanatory variable.

The average value vector of Y is given by the following equation (4)
As shown in 8),

(Equation 26) It is the average value of the objective variable Y.

The algorithm of the PLS1 analysis in the above case is shown below.

Step 1 Difference matrix X. Is calculated by the following equation (49).

[Equation 27]

Step 2 Difference matrix Y. Is calculated by the following equation (50).

[Equation 28]

Step 3 It is assumed that the number of components a = 1.

[0110] The Step 4 following equation (51) calculating the PLS weight vector W _a.

(Equation 29)

Step 5 The latent variable vector _Ta is calculated by the following equation (52). T _a = X _a-1 W _a (52)

[0112] To calculate the loading vector P _a in step 6 the following equation (53). ^{_{P a = X T a-1}} T a / (T a T T a) ......... (53)

[0113] calculating the coefficients vector Q _a in step 7 the following equation (54). Q. ^{_{= Y T a-1 T a}} / (T a T T a) ......... (54)

Step 8 The explanatory variable vector _Xa is calculated by the following equation (55). _{X a = X a-1 -T} a P a T ......... (55)

[0115] calculating the objective variable vector Y _a in step 9 the following equation (56). _{Y a = Y a-1 -T} a Q a ......... (56)

Step 10 The number of components is set to a = a + 1.

Step 11 If the next PLS component is necessary, the above steps 4 and thereafter are repeated. If the next PLS component is not needed, the program ends.

PLS for building a model as described above
In the method, the fitting error (SEC) and the prediction error (SEP) change depending on the degree of freedom of the model. FIG. 3 is a diagram illustrating the relationship between the degree of freedom of the model and the error. As shown in the figure, when the degree of freedom A of the model is small, the fitting error is large.
As the degree of freedom increases, the fit error decreases. On the other hand, the prediction error is large when the degree of freedom A of the model is small. However, when the degree of freedom is increased, the prediction error becomes minimum at a certain value _Aopt , and thereafter starts to increase again. The degree of freedom of the model is
Depends on the number of PLS components. Therefore, it can be seen that there is an optimal value for the number of PLS components (the number of factors) that minimizes the error.

The optimum number of components is obtained by cross-validation. Cross-validation includes external validation and internal validation. Less than,
These will be described.

External validation is performed when a given number of samples is relatively large. The procedure of this external validation is as follows.

1) Divide the n samples into a training set and a test set.

2) One component PL based on training set data
Deriving the S model.

3) The estimated value y _pred of the objective variable y of the test set is obtained by using the one-component PLS model.

4) Assuming that the actual measured value of y is y _obs , PRES is calculated as a prediction error by the above-mentioned estimated value y _pred of y.
S (Prediction Residual Sum of Squares) is calculated by the following equation (5
Determined by 7).

[Equation 30] Here, the sum is taken for a sample of the test set.

5) Next, regarding the data of the test set, 2
Compute the component PLS model.

6) Using the two-component PLS model, obtain the estimated value y _pred of the objective variable y of the test set.

7) The PRESS based on the two-component PLS model is obtained from the equation (57).

8) By further increasing the PLS component,
Calculate SS sequentially.

9) As shown in FIG. 3, when PRESS is minimized for a certain number of PLS components, the number of components is determined as the optimum number of components.

Next, the internal validation will be described. Internal validation is often performed using the "leave-one-out method". This "leave-one-out method"
The procedure of internal validation by is as follows.

1) In the case of n samples, a test set excluding one sample is set as a test set.

2) The remaining (n-1) samples are used as a training set, and a one-component PLS model is calculated for the training set data.

3) The estimated value y _pred of the sample of one removed test set is obtained.

4) The samples to be removed are changed one after another, and y _pred of n samples is obtained.

5) The PRESS in the case of one component is expressed by the following formula (5)
Determined by 8).

(Equation 31) Here, the sum is taken for all samples.

6) In the same procedure, the estimated value y _pred of y is obtained for all the samples for the two-component PLS model.

7) The PRESS based on the two-component PLS model is obtained from equation (58).

8) By further increasing the PLS component,
Calculate SS sequentially.

9) As shown in FIG. 3, when PRESS is minimized for a certain number of PLS components, the number of components is determined as the optimum number of components.

Note that the prediction error (SEP) may not be a downwardly convex V-shaped curve as shown in FIG.
In such a case, the optimum number of components is determined by increasing the number of PLS components and examining the standard deviation and the like.

As a measure of the prediction accuracy of the PLS model, the square value R _pred ² of the predictive correlation function may be used instead of the above PRESS. R _pred ² is defined by the following equation (59).

(Equation 32)

(5) Test evaluation process By using the PLS regression analysis method and the cross-validation described above, 42 predictive model expressions could be obtained. In these prediction model equations, the absorbance at a wavelength of 808 nm is x1, and the absorbance at a wavelength of 816nm is x1.
2, the absorbance at a wavelength of 825 nm is x3, the absorbance at a wavelength of 833 nm is x4, and the wavelength is 841n.
The absorbance at 850 nm is x5, the absorbance at 850 nm is x6, and the absorbance at 858 nm is x5.
7, the absorbance at a wavelength of 866 nm is x8, the absorbance at a wavelength of 875 nm is x9, and the wavelength is 883n.
The absorbance at a wavelength of 891 nm is x11, the absorbance at a wavelength of 891 nm is x11, and the absorbance at a wavelength of 900 nm is x
12, the absorbance at a wavelength of 908 nm is x13, the absorbance at a wavelength of 917 nm is x14,
The absorbance at a wavelength of 933 nm is x16, and the absorbance at a wavelength of 942 nm is x15.
17, the absorbance at a wavelength of 950 nm is x18, the absorbance at a wavelength of 958 nm is x19, and the wavelength is 967 n
The absorbance at a wavelength of 975 nm is x21, and the absorbance at a wavelength of 983 nm is x21.
22; the absorbance at a wavelength of 992 nm is x23; the absorbance at a wavelength of 1000 nm is x24;
The absorbance at 8 nm is x25, the absorbance at 1017 nm is x26, the absorbance at 1025 nm is x27, and the absorbance at 1033 nm is x
28, the absorbance at a wavelength of 1042 nm is x29,
The absorbance at a wavelength of 1050 nm is x30,
The absorbance at 58 nm is x31, and the wavelength is 1066 nm.
And the absorbance at a wavelength of 1075 nm as x33, the amylose content y (%) in the polished rice sample is calculated by the following formula (60): y = b1 × x1 + b2 × x2 + b3 × x3 + b4 × x4 + b5 × x5 + b6 * x6 + b7 * x7 + b8 * x8 + b9 * x9 + b10 * x10 + b11 * x11 + b12 * x12 + b13 * x13 + b14 * x14 + b15 * x15 + b16 * x16 + b17 * x17 + b18 * x18 + 22 * 23 * 22 * 22 * b * 21 * 22 * 22 * b * 20 * 22 * b * 20 * 22 * b * 21 * 22 * b * 20 * 22 * b * 21 * 22 * b * 21 * 22 * b * 21 * 22 * b * 21 * 22 * b * 22 * 22 * b * 22 * b * 22 * b * 21) × x26 + b27 × x27 + b28 × x28 + b29 × x29 + b30 × x30 + b31 × x31 + b32 × x32 + b33 × x33 + b. ... (60)

The PL obtained when these prediction model equations are obtained
The number of S components (the number of factors) was 3-33. In the above equation (60), b1 to b33 are coefficients and b. Is a constant term. In the 42 prediction model equations, these coefficients b1 to b33 and the constant term b. Is in the following range. -222 ≤ b1 ≤ 1790 -544 ≤ b2 ≤ 2140 -5100 ≤ b3 ≤ 1520 -103000 ≤ b4 ≤ 1190-74700 ≤ b5 ≤ 35500 -52700 ≤ b6 ≤ 53700 -104000 ≤ b7 ≤ 5550 -80600 ≤ b8 ≤ 29100-60700 ≤ b9 ≤ 51200 -107000 ≤ b10 ≤ 5750 -83700 ≤ b11 ≤ 33400 -50600 ≤ b12 ≤ 58200 -100 000 ≤ b13 ≤ 6350-84200 ≤ b14 ≤ 24600 -60400 ≤ b15 ≤ 50700 -98800 ≤ b16 ≤ 7430 -78700 ≤ b17 ≤ 33600 -63900 ≤ b18 ≤ 50200 -95400 ≤ b19 ≤ 6070 -78000 ≤ b20 ≤ 25600 -52300 ≤ b21 ≤ 56000 -103000 ≤ b22 ≤ 2870 -78300 ≤ b23 ≤ 32400-62200 ≤ b24 ≤ 52300 -99500 ≤ b25 ≤ 3690-77800 ≤ b26 ≤ 34600 -59800 ≤ b27 ≤ 52300 -103,000 ≤ b28 ≤ 4090 -80900 ≤ b29 ≤ 30200 -56000 ≤ b30 ≤ 54300-1690 ≤ b31 ≤2980-1710≤b32≤1540-1980≤b33≤1400-201000≤b. ≤1270000

The 42 amylose prediction models were evaluated for predictability. As a result, the prediction error (SEP) of each model was 0.49 to 1.30. The multiple correlation coefficient at the time of the test is 0.519 to 0.9.
It was 41.

From this, it can be seen that these 42 prediction model equations are low-error prediction errors (SEP) and highly correlated (highly accurate) amylose content prediction equations.

Next, from the above 42 prediction model equations, the three wavelengths having the shortest wavelengths, 808 nm and 816n, are used.
Excluding m and 825 nm, and excluding the longest three wavelengths of 1058 nm, 1066 nm and 1075 nm, 19 predictive model equations using only the intermediate 27 wavelengths of 833 nm to 1050 nm were extracted.

[0147] This is because the spectral characteristics of the spectrometer are unstable near both ends of the light region to be used, and even if a good calibration curve is obtained by one device, the calibration curve is converted to another device of the same model. This is because there is a problem of "machine difference" that an error is likely to occur when the method is applied.

The number of PLS components (the number of factors) when these 19 prediction model equations were obtained was 3 to 27. In these 19 prediction model equations, the coefficient b
4 to b30 and the constant term b. Is in the following range. −26133 ≦ b4 ≦ 957−64320 ≦ b5 ≦ 25608−959 ≦ b6 ≦ 53699−21172 ≦ b7 ≦ 5550−71496 ≦ b8 ≦ 21526−3954 ≦ b9 ≦ 51172−20859 ≦ b10 ≦ 5748−72463 ≦ b11 ≦ 266103 ≦ b12≤58193-20104≤b13≤6350-73737≤b14≤18320-4475≤b15≤50700-18813≤b16≤7428-67991≤b17≤25117-6679≤b18≤50200-22797≤b19≤5056 -70343≤b20≤ 18703 33 ≤ b21 ≤ 55996-23377 ≤ b22 ≤ 2866 -68306 ≤ b23 ≤ 24348 -5034 ≤ b24 ≤ 52266 -22473 ≤ b25 ≤ 3690 -68729 ≤ b26 ≤ 26265 -399 ≦ b27 ≦ 52300 -22176 ≦ b28 ≦ 4090 -70068 ≦ b29 ≦ 21690 0 ≦ b30 ≦ 54300 -201000 ≦ b. ≤346450

For these 19 amylose prediction models, test evaluation of predictability was performed. As a result, the prediction error (SEP) of each model was 0.49 to 1.275. The multiple correlation coefficient at the time of the test is 0.53 to 0.94.
It became 1.

From this, it can be seen that these 19 prediction model formulas have a lower prediction error (SEP) and a higher correlation (higher precision) amylose content prediction formula.

Next, 13 prediction model expressions having a prediction error (SEP) of 0.79 or less were extracted from the above 19 prediction model expressions.

The number of PLS components (the number of factors) when these 13 prediction model equations were obtained was 9-12. In these 13 prediction model equations, the coefficient b
4 to b30 and the constant term b. Is in the following range. -2713 ≤ b4 ≤ -68 501 ≤ b5 ≤ 5442 -959 ≤ b6 ≤ 2667 -4059 ≤ b7 ≤ 5550-2586 ≤ b8 ≤ -658 -3416 ≤ b9 ≤ -159 -765 ≤ b10 ≤ 5158 -2894 ≤ b11 ≤ 3586 2199 ≤ b12 ≤ 4866 -215 ≤ b13 ≤ 6350 -4814 ≤ b14 ≤ 2213 -4020 ≤ b15 ≤ -2287 2439 ≤ b16 ≤ 7428 259 ≤ b17 ≤ 2271 -4561 ≤ b18 ≤ -1415-1844 ≤ b19 ≤ 3974-1722 ≤ b20 ≤ -732 1612 ≤ b21 ≤ 4287-2239 ≤ b22 ≤ 2866 665 ≤ b23 ≤ 1840 -3436 ≤ b24 ≤ -884 -680 ≤ b25 ≤ 3690 463 ≤ b26 ≤ 3438 -2176 ≤ b27 ≤ -445 -595 ≤ b28 ≤ 4090-2121 ≦ b29 ≦ 7680 ≦ b3 0 ≦ 2004-26592 ≦ b. ≤ 69

For these 13 amylose prediction models, a test evaluation of predictability was performed. As a result, the prediction error (SEP) of each model was 0.49 to 0.79. The multiple correlation coefficient at the time of the test is 0.85 to 0.94.
It became.

From this, it can be seen that these 13 prediction model equations are prediction equations of amylose having a further lower prediction error (SEP) and a higher correlation (higher accuracy).

Next, seven prediction model expressions having a prediction error (SEP) smaller than 0.78 were extracted from the above 19 prediction model expressions.

The number of PLS components (the number of factors) when these seven prediction model equations were obtained was 9-13. In these seven prediction model equations, the coefficient b4
~ B30 and constant term b. Is in the following range. -2713 ≤ b4 ≤ -67 666 ≤ b5 ≤ 2473 -237 ≤ b6 ≤ 1324 -518 ≤ b7 ≤ 835 -1340 ≤ b8 ≤ -658 -842 ≤ b9 ≤ -159 -14 ≤ b10 ≤ 675 -1157 ≤ b11 ≤ 164 2199 ≤ b12 ≤ 2624 750 ≤ b13 ≤ 2223 -2890 ≤ b14 ≤ -2692-3390 ≤ b15 ≤ -2287 2439 ≤ b16 ≤ 3340 1640 ≤ b17 ≤ 2088 -2458 ≤ b18 ≤ -1415 -1884 ≤ b19 ≤ -1406 -1533 ≤ b20 ≤ -732 3080 ≤ b21 ≤ 4287 -1639 ≤ b22 ≤ 1120 665 ≤ b23 ≤ 1446 -1120 ≤ b24 ≤ -884 -680 ≤ b25 ≤ -87 463 ≤ b26 ≤ 1279 -1264 ≤ b27 ≤ -445 -508 ≤ b28 ≦ 38 37 ≦ b29 ≦ 386 311 ≦ b30 ≦ 938-47 b. ≦ −17

For these seven amylose prediction models, test evaluation of predictability was performed. As a result, the prediction error (SEP) of each model was 0.72 to 0.77.
The multiple correlation coefficient at the time of the test was 0.86 to 0.87.

From this, it can be seen that these seven prediction model equations are prediction equations of amylose having very low prediction error (SEP) and very high correlation (very good precision). .

The present invention can be realized by other embodiments. FIG. 4 is a conceptual block diagram showing a configuration of a rice amylose content measuring device according to another embodiment of the present invention.

As shown in FIG. 4, the rice amylose content measuring device 10 comprises a light irradiating section 1, an absorbance measuring section 2,
It comprises a data analysis unit 3 and an output unit 4.

The light irradiating section 1 has a near infrared wavelength range (wavelength 808).
10 to 1075 nanometers) (not shown). As the light source, a heat light source, a light emitting diode (LED), a semiconductor laser (LD), another laser, or the like is used. The absorbance measuring section 2 has a light detector (not shown). As the photodetector, a "photoconductive effect element" such as a zinc sulfide element, a "photovoltaic effect element" such as a germanium photodiode, a multi-channel detector, and the like are used.

The sample S is disposed between the light irradiating section 1 and the absorbance measuring section 2 and receives irradiation light L from the light irradiating section 1. Is sample S
The transmitted light Lt absorbed by the light source is measured by the absorbance measuring unit 2.

A spectroscopy system (not shown) is provided either between the light irradiating section 1 and the sample S or between the sample S and the absorbance measuring section 2. As a method of spectroscopy, various methods such as a method using an interference filter, a multi-channel Fourier transform spectroscopy, an acousto-optic dispersion spectroscopy, and a spectroscopy using a monochromator (monochromatic spectroscopy) are used.

When the spectroscopic system is provided between the light irradiation unit 1 and the sample S, the near-infrared light L from the light source. Is transmitted to the sample S after being divided into light of individual wavelengths by the spectroscopic system, and the transmitted light Lt after being absorbed by the sample is measured by the absorbance measuring unit 2. In this case, the spectrum is obtained by wavelength sweeping the spectroscopy system.

When the spectroscopic system is provided between the sample S and the absorbance measuring section 2, the near-infrared light L from the light source. But,
The transmitted light L irradiated to the sample S and absorbed by the sample S
t is directed to the spectroscopy system, which obtains a spectrum.

The data analysis section 3 is constituted by a microcomputer (not shown). This microcomputer has a CPU (Central Processing Unit) (not shown).
nit: central processing unit) and ROM (not shown)
Only Memory: Read-only memory) and RA not shown
M (Random Access Memory)
And an input operation unit (not shown).

The CPU controls the light irradiation unit 1 and the absorbance measurement unit 2 and calculates and analyzes data. ROM
Is a storage device that stores programs executed by the CPU. The RAM is a storage device that temporarily stores data and the like calculated by the CPU. The input operation unit is C
This is a part for giving a command to the PU or inputting a program, data, or the like to a ROM or the like. The output unit 4 includes the data analysis unit 3
This is a section for outputting an amylose content value or the like output from the printer as an image, a printout, or the like.

The absorbance measuring section 2 outputs the measured absorption spectrum to the data analyzing section 3. CP of data analysis unit 3
U calculates the absorbance xi at 33 wavelengths from the absorption spectrum. Light intensity I of near-infrared light of each wavelength before transmission through the sample. And the logarithm of the common logarithm is the light intensity It after transmission through the sample, and the absorbance xi is calculated by the following equation (72): xi = log (I./It) (72) The absorbances xi at the respective wavelengths are the same as in the above-described method for measuring the amylose content of rice.

The above-mentioned amylose content prediction formula (for example, the above formula (60)) is stored in the ROM of the data analysis unit 3. When the CPU calculates the absorbance xi,
The amylose content value is calculated by the prediction formula in the ROM and output to the output unit 4. The output unit 4 includes an image display (not shown) or a printer (not shown), and outputs the measured amylose content value (%) in the sample S as an image or a print.

As described above, the amylose content prediction model stored in the data analysis unit 3 of this apparatus can accurately predict the amylose content with a very low prediction error as compared with the conventional model. With this device, the same effect can be obtained.

The present invention is not limited to the above embodiments. Each of the above embodiments is merely an example, and has substantially the same configuration as the technical idea described in the claims of the present invention, and those having the same functions and effects are:
Anything is included in the technical scope of the present invention.

For example, in the above-mentioned prediction model equation (60), the absorbance at a wavelength of 808 nm is x1, but the present invention is not limited to this, and it is not limited to x1.
The absorbance at an arbitrary wavelength within the range (808 ± 4 nm) may be defined as x1. Further, in the above-described prediction model formula (60), the absorbance at a wavelength of 816 nm is x2, but the present invention is not limited to this, and
The absorbance at an arbitrary wavelength within the range of 820 nm (816 ± 4 nm) may be defined as x2. Further, in the above-mentioned prediction model formula (60), the absorbance at a wavelength of 825 nm is x3, but the present invention is not limited to this.
The absorbance at an arbitrary wavelength within the range of 821 to 829 nm (825 ± 4 nm) may be defined as x3. In the above-described prediction model equation (60), the wavelength 833 nm
Is x4, but the present invention is not limited to this, and the absorbance at any wavelength in the range of 829 to 837 nm (833 ± 4 nm) may be x4. In the above-described prediction model equation (60), the wavelength 8
Although the absorbance at 41 nm was defined as x5, the present invention is not limited thereto, and the absorbance at 838 to 845 nm (841 ± 4n
The absorbance at any wavelength within the range of m) may be x5. Further, in the above-mentioned prediction model equation (60), the absorbance at a wavelength of 850 nm is x6, but the present invention is not limited to this, and the present invention is not limited to this.
0 ± 4 nm) at any wavelength within the range x
It may be 6. Further, the above-described prediction model equation (60)
In the above, the absorbance at a wavelength of 858 nm was x7, but the present invention is not limited to this, and 854 to 862 n
The absorbance at any wavelength within the range of m (858 ± 4 nm) may be x7. Further, in the above-mentioned prediction model formula (60), the absorbance at a wavelength of 866 nm is set to x8, but the present invention is not limited to this, and
The absorbance at any wavelength within the range of 870 nm (866 ± 4 nm) may be x8. Further, in the above-mentioned prediction model formula (60), the absorbance at a wavelength of 875 nm is x9, but the present invention is not limited to this.
The absorbance at an arbitrary wavelength within the range of 871 to 879 nm (875 ± 4 nm) may be defined as x9. In the above-mentioned prediction model equation (60), the wavelength 883 nm
Is x10, but the present invention is not limited to this, and the absorbance at any wavelength in the range of 879-887 nm (883 ± 4 nm) may be x10. In the above-described prediction model equation (60), the wavelength 8
Although the absorbance at 91 nm was defined as x11, the present invention is not limited to this, and 887 to 895 nm (891 ± 4n
The absorbance at any wavelength within the range of m) may be x11. Further, in the above-mentioned prediction model formula (60), the absorbance at a wavelength of 900 nm is x12, but the present invention is not limited to this, and it is not limited to x12.
0 ± 4 nm) at any wavelength within the range x
It may be set to 12. Further, the above-described prediction model equation (60)
In the above, the absorbance at a wavelength of 908 nm was x13, but the present invention is not limited to this, and 904 to 912n
The absorbance at an arbitrary wavelength within the range of m (908 ± 4 nm) may be x13. Further, in the above-described prediction model equation (60), the absorbance at a wavelength of 917 nm is x14, but the present invention is not limited to this, and the present invention is not limited thereto.
The absorbance at any wavelength within the range of 921 nm (917 ± 4 nm) may be x14. In the above-mentioned prediction model formula (60), the absorbance at a wavelength of 925 nm is x15, but the present invention is not limited to this.
The absorbance at any wavelength within the range of 921 to 929 nm (925 ± 4 nm) may be x15. In the above-mentioned prediction model equation (60), the wavelength 933 nm
Is x16, but the present invention is not limited to this, and the absorbance at any wavelength in the range of 929 to 937 nm (933 ± 4 nm) may be x16. In the above-described prediction model equation (60), the wavelength 9
Although the absorbance at 42 nm was x17, the present invention is not limited to this, and 938 to 946 nm (942 ± 4n
The absorbance at an arbitrary wavelength within the range of m) may be x17. Further, in the above-described prediction model formula (60), the absorbance at a wavelength of 950 nm is x18, but the present invention is not limited to this, and it is not limited to x18.
0 ± 4 nm) at any wavelength within the range x
It may be 18. Further, the above-described prediction model equation (60)
In the above, the absorbance at a wavelength of 958 nm was set to x19, but the present invention is not limited to this, and 954 to 962n
The absorbance at an arbitrary wavelength within the range of m (958 ± 4 nm) may be x19. In the above-described prediction model formula (60), the absorbance at a wavelength of 967 nm is x20, but the present invention is not limited to this, and
The absorbance at any wavelength within the range of 971 nm (967 ± 4 nm) may be x20. In the above-mentioned prediction model formula (60), the absorbance at a wavelength of 975 nm is x21, but the present invention is not limited to this.
The absorbance at an arbitrary wavelength within the range of 971 to 979 nm (975 ± 4 nm) may be x21. In the above-mentioned prediction model equation (60), the wavelength 983 nm
Is x22, but the present invention is not limited to this, and the absorbance at any wavelength in the range of 979 to 987 nm (983 ± 4 nm) may be x22. In the above-described prediction model equation (60), the wavelength 9
Although the absorbance at 92 nm was x23, the present invention is not limited to this, and the absorbance is 988 to 996 nm (992 ± 4n
The absorbance at any wavelength within the range of m) may be x23. In the above-mentioned prediction model formula (60), the absorbance at a wavelength of 1000 nm is x24.
The present invention is not limited to this,
The absorbance at any wavelength within the range (1000 ± 4 nm) may be x24. Further, in the above-mentioned prediction model formula (60), the absorbance at a wavelength of 1008 nm is x25, but the present invention is not limited to this, and is not limited to x25.
The absorbance at any wavelength within the range of 4 to 1012 nm (1008 ± 4 nm) may be x25. In the above-mentioned prediction model equation (60), the wavelength 1017n
Although the absorbance at m is x26, the present invention is not limited to this, but is 1013 to 1021 nm (1017 ± 4n
The absorbance at any wavelength within the range of m) may be x26. In the above-mentioned prediction model formula (60), the absorbance at a wavelength of 1025 nm is x27.
The present invention is not limited to this, and 1021 to 1029 nm
The absorbance at any wavelength within the range (1025 ± 4 nm) may be x27. In the above-described prediction model formula (60), the absorbance at a wavelength of 1033 nm is x28, but the present invention is not limited to this, and
The absorbance at an arbitrary wavelength within the range of 9 to 1037 nm (1033 ± 4 nm) may be defined as x28. In the above-described prediction model equation (60), the wavelength 1042n
Although the absorbance at m is x29, the present invention is not limited to this, but is 1038 to 1046 nm (1042 ± 4n
The absorbance at any wavelength within the range of m) may be x29. In the above-mentioned prediction model formula (60), the absorbance at a wavelength of 1050 nm is x30.
The present invention is not limited to this, but 1046 to 1054 nm
The absorbance at any wavelength within the range (1050 ± 4 nm) may be x30. In the above-described prediction model formula (60), the absorbance at a wavelength of 1058 nm is x31, but the present invention is not limited to this, and
The absorbance at any wavelength within the range of 4 to 1062 nm (1058 ± 4 nm) may be defined as x31. In the above-described prediction model equation (60), the wavelength 1066n
Although the absorbance at m is x32, the present invention is not limited to this, and the absorbance at 1062 to 1070 nm (1066 ± 4n
The absorbance at any wavelength within the range of m) may be x32. In the above-mentioned prediction model formula (60), the absorbance at a wavelength of 1075 nm is x33,
The present invention is not limited to this, and 1071 to 1079 nm
The absorbance at an arbitrary wavelength within the range of (1075 ± 4 nm) may be x33. Even if the wavelength of the near-infrared light to be used is within these ranges, an accurate prediction model formula can be obtained. If the wavelength of the near-infrared light to be used is within these ranges, the entire wavelength range of the near-infrared light is 804
1079 nm.

[0173]

As described above, according to the present invention,
As compared with the conventional method, there is an advantage that the amylose content can be accurately predicted with a very low prediction error.

[Brief description of the drawings]

FIG. 1 is a diagram showing a distribution of amylose content of a milled rice sample of general rice collected in a sample collecting process of a method for measuring amylose content of rice according to an embodiment of the present invention.

FIG. 2 is a diagram showing an example of an absorption spectrum of a polished rice sample of general rice for use in a method for measuring amylose content of rice according to an embodiment of the present invention, which is measured in a near-infrared spectroscopic measurement process of an amylose content value.

FIG. 3 shows a degree of freedom A of a model in PLS modeling.
FIG. 6 is a diagram showing a relationship between a matching error and a prediction error for

FIG. 4 is a conceptual block diagram illustrating a configuration of an apparatus for measuring amylose content of rice according to another embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Light irradiation part 2 Absorbance measurement part 3 Data analysis part 4 Output part 10 Amylose content measuring device L of rice. Irradiation light Lt Transmitted light S Sample

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] April 9, 1999

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0012

[Correction method] Change

[Correction contents]

[0012]

Means for Solving the Problems] To solve the above problems, the amylose content measuring method rice according to the present invention, in
Japanese rice other than Do rice and Java rice
From general rice for rice in various parts of Japan from Japan to Kyushu, including amylose
Amount distribution is 18-23%, from semi-mochi rice to high amylose
Reduce the amount of polished rice samples including those of multiple varieties up to rice
And 110 samples were collected and the milled rice yield of the milled rice sample
Of the rice obtained from the polished rice sample.
Colorimetric determination after adding iodine reagent to pung gelatinization liquid
The case was unified to a temperature of 25 ° C and a time of 20 minutes.
Base content reference value
Irradiates near-infrared light with a wavelength of 804 to 1079 nanometers
The near-infrared light is reflected by the polished rice sample at each wavelength.
Amylose content value to measure absorbance absorbed
Perform a spectroscopic measurement process, the amylose content reference value chemical
Practice gained by chemical analysis methods during the measurement process
The amylose content value of
Near-infrared spectroscopy in the near-infrared spectroscopy
Partial least squares regression using measured absorbance as an explanatory variable
By performing the analysis, the amylose in the polished rice sample
The prediction formula for predicting the content y (%) is expressed by wavelength 808
The absorbance at ± 4 nm is x1, and the wavelength is 816 ± 4
The absorbance at nm is x2 and the wavelength is 825 ± 4 nm
X3, the absorbance at 833 ± 4 nm is x4, the absorbance at 841 ± 4 nm is x5, the absorbance at 850 ± 4 nm is x6, the absorbance at 858 ± 4 nm is x7, and the wavelength is 866 ± 4 nm. The absorbance at x8, the absorbance at a wavelength of 875 ± 4 nm as x9,
The absorbance at a wavelength of 883 ± 4 nm is x10, the absorbance at a wavelength of 891 ± 4 nm is x11, and the wavelength is 9
The absorbance at 00 ± 4 nm is x12 and the wavelength is 908
The absorbance at ± 4 nm is x13 and the wavelength is 917 ± 4
The absorbance at nm is x14 and the wavelength is 925 ± 4 nm
The absorbance at wavelength 933 ± 4 nm is x16, the absorbance at 942 ± 4 nm is x17, the absorbance at 950 ± 4 nm is x18, the absorbance at 958 ± 4 nm is x19, and the wavelength is 967 ± 4 nm. Absorbance at x20
And the absorbance at a wavelength of 975 ± 4 nm is x21, the absorbance at a wavelength of 983 ± 4 nm is x22,
The absorbance at a wavelength of 992 ± 4 nm is x23, the absorbance at a wavelength of 1000 ± 4 nm is x24, and the wavelength is 10
The absorbance at 08 ± 4 nm is x25, and the wavelength is 1017.
The absorbance at ± 4 nm is x26, and the wavelength is 1025 ± 4.
The absorbance at nm is x27 and the wavelength is 1033 ± 4 nm.
, The absorbance at a wavelength of 1042 ± 4 nm is x29, the absorbance at a wavelength of 1050 ± 4 nm is x30, the absorbance at a wavelength of 1058 ± 4 nm is x31, the absorbance at a wavelength of 1066 ± 4 nm is x32, and the wavelength is 1075 ± 4 nm. Absorbance at x33
Where b1 = b2 = b3 = b31 = b32 = b3
3 = 0, and the coefficient b4 is set to −2713 ≦ b4 ≦ −
67 and the coefficient b5 is 666 ≦ b5 ≦ 247
3, and the coefficient b6 is -237 ≦ b6 ≦ 1324
And the coefficient b7 is -518 ≦ b7 ≦ 835.
And the coefficient b8 is expressed as −1340 ≦ b8 ≦ −658.
And the coefficient b9 is -842 ≦ b9 ≦ −159.
And the coefficient b10 is set to −14 ≦ b10 ≦ 675.
And the coefficient b11 is defined as −1157 ≦ b11 ≦ 164.
And the coefficient b12 is 2199 ≦ b12 ≦ 2624.
And the coefficient b13 is defined as 750 ≦ b13 ≦ 2223.
And the coefficient b14 is -2890 ≦ b14 ≦ −2692.
And the coefficient b15 is −3390 ≦ b15 ≦ −2287.
And the coefficient b16 is defined as 2439 ≦ b16 ≦ 3340.
And the coefficient b17 is 1640 ≦ b17 ≦ 2088
And the coefficient b18 is set to −2458 ≦ b18 ≦ −1415.
And the coefficient b19 is -1884≤b19≤-1406.
And the coefficient b20 is set to −1533 ≦ b20 ≦ −732.
And the coefficient b21 is defined as 3080 ≦ b21 ≦ 4287.
And the coefficient b22 is set to -1639 ≦ b22 ≦ 1120.
And the coefficient b23 is set to 665 ≦ b23 ≦ 1446.
And the coefficient b24 is defined as -1120 ≦ b24 ≦ −884.
And the coefficient b25 is -680 ≦ b25 ≦ −87
And the coefficient b26 is 463 ≦ b26 ≦ 1279.
And the coefficient b27 is set to −1264 ≦ b27 ≦ −445.
And the coefficient b28 is -508 ≦ b28 ≦ 38
And the coefficient b29 is expressed as 37 ≦ b29 ≦ 386.
And the coefficient b30 is 311 ≦ b30 ≦ 938
And the constant term b. −47 ≦ b. ≤ -17
And y = b1.times.x1 + b2.times.x2 + b3.times.x3 + b4.times.x4 + b5.times.x5 + b6.times.x6 + b7.times.x7 + b8.times.x8 + b9.times.x9 + b10.times.x10 + b11.times.11 + b12.times.x12 + b13.times.x13 + b14.times.x16 + b15.times.x16 + b15.times. + B21 * x21 + b22 * x22 + b23 * x23 + b24 * x24 + b25 * x25 + b26 * x26 + b27 * x27 + b28 * x28 + b29 * x29 + b30 * x30 + b31 * x31 + b32 * x32 + b33 * x33 + b. It is characterized by being calculated by:

[Procedure amendment 3]

[Document name to be amended] Statement

[Correction target item name] 0013

[Correction method] Change

[Correction contents]

Further, the amylose content of rice according to the present invention is measured.
The setting device is for Japanese rice other than Indian rice and Java rice.
Oh, from Hokkaido to Kyushu
The amylose content distribution becomes 18-23%,
A variety of varieties ranging from rice to high amylose rice
Collecting at least 110 samples of white rice,
Adjust the milled rice yield of rice sample to 90 ± 1%,
After adding iodine reagent to starch gelatinization solution obtained from sample
Conditions for colorimetric determination at 25 ° C for 20 minutes
Amylose content reference value chemical measurement process
The polished rice sample has a wavelength of 804 to 1079 nanometers.
The near-infrared light at each wavelength
An amino acid measuring absorbance absorbed by the milled rice sample
Perform the near-infrared spectroscopic measurement process of the loin content value, and
Standard value for chemical analysis in chemical measurement process
Therefore, the actual amylose content value obtained was used as the target variable.
In the near-infrared spectroscopic measurement process of the amylose content value,
The absorbance measured by near-infrared spectroscopy
Predictions obtained by performing partial least squares regression analysis
Using the formula, the near-infrared wavelength of 804 to 1079 nm
A light irradiator for irradiating light to the polished rice sample, and the near infrared light
Absorbance absorbed by the milled rice sample at each wavelength
An absorbance measuring unit for measuring the amount of amylo in the milled rice sample
The above formula for predicting the source content y (%)
The absorbance at 808 ± 4 nm is x1, and the wavelength 8
The absorbance at 16 ± 4 nm is x2, and the wavelength is 825.
The absorbance at ± 4 nm is x3, and the wavelength is 833 ± 4.
The absorbance at nm is x4 and the wavelength is 841 ± 4 nm
X5 at 850 ± 4 nm
X6 at the wavelength of 858 ± 4 nm
Absorbance is x7, absorbance at wavelength 866 ± 4nm
And the absorbance at a wavelength of 875 ± 4 nm
x9 and absorbance at wavelength 883 ± 4 nm x10
And the absorbance at a wavelength of 891 ± 4 nm is x11.
And the absorbance at a wavelength of 900 ± 4 nm is x12,
The absorbance at 908 ± 4 nm is x13 and the wavelength is
The absorbance at 917 ± 4 nm is x14, and the wavelength 9
The absorbance at 25 ± 4 nm is x15, and the wavelength is 933.
The absorbance at ± 4 nm is x16 and the wavelength is 942 ± 4
The absorbance at nm is x17 and the wavelength is 950 ± 4 nm
The absorbance at x18 and the wavelength at 958 ± 4 nm
X19 at the wavelength of 967 ± 4 nm
Absorbance x20, absorbance at wavelength 975 ± 4nm
The degree of absorption is x21, and the absorbance at a wavelength of 983 ± 4 nm is
x22, and the absorbance at a wavelength of 992 ± 4 nm is x23.
And the absorbance at a wavelength of 1000 ± 4 nm is x24.
And the absorbance at a wavelength of 1008 ± 4 nm is x25,
The absorbance at a wavelength of 1017 ± 4 nm is x26, and the wavelength is
The absorbance at 1025 ± 4 nm is x27 and the wavelength is 10
The absorbance at 33 ± 4 nm is x28, and the wavelength is 1042.
The absorbance at ± 4 nm is x29 and the wavelength is 1050 ± 4
The absorbance at nm is x30, and the wavelength is 1058 ± 4 nm.
Is the absorbance at x31 and the wavelength is 1066 ± 4 nm.
X32 at the wavelength of 1075 ± 4 nm
When the absorbance is x33, the coefficient b1 = b2 = b3 = b3
1 = b32 = b33 = 0, and the coefficient b4 is set to −2713 ≦
b4 ≦ −67, and the coefficient b5 is 666 ≦ b
5 ≦ 2473, and the coefficient b6 is −237 ≦ b6
≦ 1324, and the coefficient b7 is −518 ≦ b7 ≦
835, and the coefficient b8 is −1340 ≦ b8 ≦
−658, and the coefficient b9 is −842 ≦ b9 ≦ −
159 and the coefficient b10 is −14 ≦ b10 ≦ 6
75 and the coefficient b11 is −1157 ≦ b11 ≦ 16
4 and the coefficient b12 is 2199 ≦ b12 ≦ 2624
And the coefficient b13 is 750 ≦ b13 ≦ 2223.
And the coefficient b14 is -2890 ≦ b14 ≦ −2692.
And the coefficient b15 is −3390 ≦ b15 ≦ −2287.
And the coefficient b16 is defined as 2439 ≦ b16 ≦ 3340.
And the coefficient b17 is 1640 ≦ b17 ≦ 2088
And the coefficient b18 is set to −2458 ≦ b18 ≦ −1415.
And the coefficient b19 is -1884≤b19≤-1406.
And the coefficient b20 is set to −1533 ≦ b20 ≦ −732.
And the coefficient b21 is defined as 3080 ≦ b21 ≦ 4287.
And the coefficient b22 is set to -1639 ≦ b22 ≦ 1120.
And the coefficient b23 is set to 665 ≦ b23 ≦ 1446.
And the coefficient b24 is defined as -1120 ≦ b24 ≦ −884.
And the coefficient b25 is -680 ≦ b25 ≦ −87
And the coefficient b26 is 463 ≦ b26 ≦ 1279.
And the coefficient b27 is set to −1264 ≦ b27 ≦ −445.
And the coefficient b28 is -508 ≦ b28 ≦ 38
And the coefficient b29 is expressed as 37 ≦ b29 ≦ 386.
And the coefficient b30 is 311 ≦ b30 ≦ 938
And the constant term b. −47 ≦ b. ≤ -17
The following equation is used: y = b1.times.x1 + b2.times.x2 + b3.times.x3 + b4.times.b4.times.b5.times.x5 + b6.times.x6 + b7.times.x7 + b8.times.x8 + b9.times.x9 + b10.times.x10 + b11.times.11 + b12.times.x12 + b13.times.x13 + b14.times.x16 + b15.times.x16 + b15.times. + B21 * x21 + b22 * x22 + b23 * x23 + b24 * x24 + b25 * x25 + b26 * x26 + b27 * x27 + b28 * x28 + b29 * x29 + b30 * x30 + b31 * x31 + b32 * x32 + b33 * x33 + b. And a data analysis unit that calculates the following.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0014

[Correction method] Deleted

[Procedure amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0015

[Correction method] Deleted

[Procedure amendment 6]

[Document name to be amended] Statement

[Correction target item name] 0016

[Correction method] Deleted

[Procedure amendment 7]

[Document name to be amended] Statement

[Correction target item name] 0017

[Correction method] Deleted

[Procedure amendment 8]

[Document name to be amended] Statement

[Correction target item name] 0018

[Correction method] Deleted

[Procedure amendment 9]

[Document name to be amended] Statement

[Correction target item name] 0019

[Correction method] Deleted

 ──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Takeshi Yanagisawa 3-17-7 Okubo, Konan-ku, Yokohama-shi, Kanagawa Prefecture (72) Inventor Bunji Tezuka 5-27-15-702 Oyada, Adachi-ku, Tokyo (72) Inventor Yasuhiro Maruyama 202-11 Cosmo Sanno 1-11-15 Sanno, Ota-ku, Tokyo Japan (72) Junji Katsura 2-3-2 Maison Abe 2-3-2 Kusashita, Soka-shi, Saitama 205 (72) Inventor Hidechika Toshima Azuma, Tsukuba-shi, Ibaraki 2-chome 13-1-909- 402 (72) Inventor Hiroshi Okadome 2-11-804-402 (72) Inventor Kenichi Otsubo 4227-10 Ami, Ami-cho, Inashiki-gun, Ibaraki Pref.

Claims (8)

[Claims] 1. A sample of polished rice is irradiated with near-infrared light having a wavelength of 804 to 1079 nm, and the absorbance at which the near-infrared light is absorbed by the polished rice sample at each wavelength is measured. Let x1 be the absorbance at
The absorbance at a wavelength of 816 ± 4 nm is x2, the absorbance at a wavelength of 825 ± 4 nm is x3,
The absorbance at 33 ± 4 nm is x4 and the wavelength is 841
The absorbance at ± 4 nm is x5 and the wavelength is 850 ± 4
The absorbance at nm is x6 and the wavelength is 858 ± 4 nm
The absorbance at 866 ± 4 nm is x8, the absorbance at 875 ± 4 nm is x9, the absorbance at 883 ± 4 nm is x10, the absorbance at 891 ± 4 nm is x11, and the wavelength is 900 ± 4 nm. Absorbance at x12
And the absorbance at a wavelength of 908 ± 4 nm is x13, the absorbance at a wavelength of 917 ± 4 nm is x14,
The absorbance at a wavelength of 925 ± 4 nm is x15, and the absorbance at a wavelength of 933 ± 4 nm is x16.
The absorbance at 42 ± 4 nm is x17, and the wavelength is 950.
The absorbance at ± 4 nm is x18, and the wavelength is 958 ± 4
The absorbance at nm is x19 and the wavelength is 967 ± 4 nm
Is x20, the absorbance at 975 ± 4 nm is x21, the absorbance at 983 ± 4 nm is x22, the absorbance at 992 ± 4 nm is x23, the absorbance at 1000 ± 4 nm is x24, and the wavelength is 1008 ± 4 nm. Absorbance at x25
And the absorbance at a wavelength of 1017 ± 4 nm is x26, the absorbance at a wavelength of 1025 ± 4 nm is x27,
The absorbance at a wavelength of 1033 ± 4 nm is x28, the absorbance at a wavelength of 1042 ± 4 nm is x29,
The absorbance at 50 ± 4 nm is x30, and the wavelength is 1058.
The absorbance at ± 4 nm is x31, and the wavelength is 1066 ± 4.
The absorbance at nm is x32, and the wavelength is 1075 ± 4 nm.
And the absorbance x1 to x33 as explanatory variables, and the amylose content y in the polished rice sample
By performing partial least squares regression analysis using (%) as the objective variable, the coefficient b1 is set to -222 ≦ b1 ≦
1790, and the coefficient b2 is -544 ≦ b2 ≦
2140, and the coefficient b3 is −5100 ≦ b3
≤ 1520 and the coefficient b4 is -103000 ≤
b4 ≦ 1190, and the coefficient b5 is −7470.
0 ≦ b5 ≦ 35500, and the coefficient b6 is −52
700 ≦ b6 ≦ 53700, and the coefficient b7 is −1
44000 ≦ b7 ≦ 5550, and the coefficient b8 is
-80600 ≦ b8 ≦ 29100, coefficient b9
Is set to −60700 ≦ b9 ≦ 51200, and the coefficient b10 is set to −107000 ≦ b10 ≦ 5750,
The coefficient b11 is −83700 ≦ b11 ≦ 33400, and the coefficient b12 is −50600 ≦ b12 ≦ 5820
0 and the coefficient b13 is -100,000 ≦ b13 ≦ 6
350 and the coefficient b14 is −84200 ≦ b14 ≦
24600, and the coefficient b15 is −60400 ≦ b15 ≦
50700, and the coefficient b16 is -98800 ≦ b
16 ≦ 7430, and the coefficient b17 is −78700
≦ b17 ≦ 33600, and the coefficient b18 is −639.
00 ≦ b18 ≦ 50200, and the coefficient b19 is −9.
5400 ≦ b19 ≦ 6070 and the coefficient b20
−78000 ≦ b20 ≦ 25600, coefficient b21 −52300 ≦ b21 ≦ 56000, coefficient b
22 is −103000 ≦ b22 ≦ 2870, coefficient b23 is −78300 ≦ b23 ≦ 32400, and coefficient b24 is −62200 ≦ b24 ≦ 5230.
0 and the coefficient b25 is -99500 ≦ b25 ≦ 3
690 and the coefficient b26 is −77800 ≦ b26 ≦
34600, and the coefficient b27 is -59800 ≦ b27 ≦
52300, and the coefficient b28 is −103000 ≦ b
28 ≦ 4090, and the coefficient b29 is −80900
≤ b29 ≤ 30200 and the coefficient b30 is -560
00 ≦ b30 ≦ 54300, and the coefficient b31 is −
1690 ≦ b31 ≦ 2980, and the coefficient b32 is
-1710≤b32≤1540, coefficient b33 is -1980≤b33≤1400, and constant term b. −201000 ≦ b. ≦ 1270000,
The amylose content y (%) in the polished rice sample was calculated by the following formula: y = b1 × x1 + b2 × x2 + b3 × x3 + b4 × x4 + b5 × x5 + b6 × x6 + b7 × x7 + b8 × x8 + b9 × x9 + b10 × x10 + b11 × x11 + b12 * x12 + b13 * x13 + b14 * x15 + b16 * x16 + b17 * x17 + b18 * x18 + b19 * x19 + b20 * x20 + b21 * x21 + b22 * x22 + b23 * x23 + b24 * x24 + b25 * x32 + 27 * 30 * 27 + b * 32 * 30 + b * 32 * b * x32 + b * 2 * x2 + b * 2 * x3 + b * 2 * x3 + b * 2 * x3 + b * 2 * x2 + b * 2 * x2 + b * 2 * x3 + bx * x2 + bx * x3 + bx2x3xbx3xbxx2xbxx2xbx2xbx3xbx2xbx3xbx3xbx3xbxbx2xbx3xbx as as shown in FIG. . A method for measuring the amylose content of rice, characterized by calculating the following. 2. The method according to claim 1, wherein said coefficient b1 = b2 = b3 = b31.
= B32 = b33 = 0, and the coefficient b4 is -2613.
3 ≦ b4 ≦ 957, and the coefficient b5 is -6
4320 ≦ b5 ≦ 25608, and the coefficient b6 is
−959 ≦ b6 ≦ 53699, and the coefficient b
7 is set to −21172 ≦ b7 ≦ 5550, the coefficient b8 is set to −71496 ≦ b8 ≦ 21526, and the coefficient b9 is set to −3954 ≦ b9 ≦ 511.
The coefficient b10 is set to -20859 ≦ b10 ≦
5748, and the coefficient b11 is -72463 ≦ b
11 ≦ 26610, and the coefficient b12 is
3 ≦ b12 ≦ 58193, and the coefficient b13 is −2.
0104 ≦ b13 ≦ 6350, and the coefficient b14 is
−73737 ≦ b14 ≦ 18320, and the coefficient b
15 is set to −4475 ≦ b15 ≦ 50700, the coefficient b16 is set to −18813 ≦ b16 ≦ 7428, and the coefficient b17 is set to −67999 ≦ b17 ≦ 251.
17 and the coefficient b18 is −6790 ≦ b18 ≦
50200, and the coefficient b19 is set to −22797 ≦ b
19 ≦ 5056, and the coefficient b20 is −7034
3 ≦ b20 ≦ 18703, and the coefficient b21 is
33 ≦ b21 ≦ 55996, and the coefficient b22 is
−23377 ≦ b22 ≦ 2866, and the coefficient b
23 is −68306 ≦ b23 ≦ 24348, the coefficient b24 is −5034 ≦ b24 ≦ 52266, and the coefficient b25 is −22473 ≦ b25 ≦ 36.
90, and the coefficient b26 is -68729 ≦ b26 ≦
26265, and the coefficient b27 is -3992 ≦ b
27 ≦ 52300, and the coefficient b28 is −2217
6 ≦ b28 ≦ 4090, and the coefficient b29 is −7
0068 ≦ b29 ≦ 21690, and the coefficient b30 is
0 ≦ b30 ≦ 54300, and the constant term b. −201000 ≦ b. ≦ 346450, a method for measuring amylose content of rice. 3. The method for measuring amylose content of rice according to claim 2, wherein the coefficients b1 = b2 = b3 = b31.
= B32 = b33 = 0, and the coefficient b4 is -2713
≦ b4 ≦ −68, and the coefficient b5 is 50
1 ≦ b5 ≦ 5442, and the coefficient b6 is -9
59 ≦ b6 ≦ 2667, and the coefficient b7 is -4
059 ≦ b7 ≦ 5550, and the coefficient b8 is −
2586 ≦ b8 ≦ −658, and the coefficient b9 is
−3416 ≦ b9 ≦ −159, and the coefficient b10 is
−765 ≦ b10 ≦ 5158, the coefficient b11 is −2894 ≦ b11 ≦ 3586, and the coefficient b12 is
2199 ≦ b12 ≦ 4866, and the coefficient b
13 is −215 ≦ b13 ≦ 6350, the coefficient b14 is −4814 ≦ b14 ≦ 2213, the coefficient b15 is −4020 ≦ b15 ≦ −2287, the coefficient b16 is 2439 ≦ b16 ≦ 7428, and the coefficient b17 is 259 ≦ b17 ≦ 2271,
The coefficient b18 is set to −4561 ≦ b18 ≦ −1415, and the coefficient b19 is set to −1884 ≦ b19 ≦ 3974.
And the coefficient b20 is −1722 ≦ b20 ≦ −73
2 and the coefficient b21 is 1612 ≦ b21 ≦ 42
87 and the coefficient b22 is −2239 ≦ b22 ≦ 2
866, and the coefficient b23 is 665 ≦ b23 ≦
1840, and the coefficient b24 is −3436 ≦ b24 ≦
−884, and the coefficient b25 is −680 ≦ b25
≦ 3690, and the coefficient b26 is 463 ≦ b
26 ≦ 3438, and the coefficient b27 is −2176 ≦
b27 ≦ −445, and the coefficient b28 is −595
≤ b28 ≤ 4090, and the coefficient b29 is -212
1 ≦ b29 ≦ 768, and the coefficient b30 is
0 ≦ b30 ≦ 2004, and the constant term b. To -26
592 ≦ b. ≦ 69, a method for measuring amylose content of rice. 4. The method for measuring amylose content of rice according to claim 3, wherein the coefficient b1 = b2 = b3 = b31.
= B32 = b33 = 0, and the coefficient b4 is -2713≤b
4 ≦ −67, and the coefficient b5 is 666 ≦ b5
≦ 2473, and the coefficient b6 is −237 ≦ b6 ≦
1324, and the coefficient b7 is -518 ≦ b7 ≦
835, and the coefficient b8 is −1340 ≦ b8 ≦ −
658, and the coefficient b9 is -842≤b9≤-1.
59 and the coefficient b10 is −14 ≦ b10 ≦ 67
5 and the coefficient b11 is −1157 ≦ b11 ≦ 164.
The coefficient b12 is 2199 ≦ b12 ≦ 2624, the coefficient b13 is 750 ≦ b13 ≦ 2223, the coefficient b14 is −2890 ≦ b14 ≦ −2692, and the coefficient b15 is −3390 ≦ b15 ≦ −2287, The coefficient b16 is set to 2439 ≦ b16 ≦ 3340, the coefficient b17 is set to 1640 ≦ b17 ≦ 2088, the coefficient b18 is set to −2458 ≦ b18 ≦ -1415, the coefficient b19 is set to −1884 ≦ b19 ≦ -1406, and the coefficient b20 is −1533 ≦ b20 ≦ −732, the coefficient b21 is 3080 ≦ b21 ≦ 4287, the coefficient b22 is -1639 ≦ b22 ≦ 1120, the coefficient b23 is 665 ≦ b23 ≦ 1446, and the coefficient b24 is − 1120 ≦ b24 ≦ −884, the coefficient b25 is −680 ≦ b25 ≦ −87, and the coefficient b26 is 463 ≦ b26 ≦ 1279 The coefficient b27 is set to −1264 ≦ b27 ≦ −445, the coefficient b28 is set to −508 ≦ b28 ≦ 38, the coefficient b29 is set to 37 ≦ b29 ≦ 386, and the coefficient b30 is set to 311 ≦ b30 ≦ 938. Constant term b. −47 ≦ b. ≦ −17, a method for measuring amylose content of rice. 5. A light irradiator for irradiating near-infrared light having a wavelength of 804 to 1079 nanometers on a polished rice sample, and an absorbance for measuring an absorbance of the near-infrared light absorbed by the polished rice sample at each wavelength. Measuring unit and wavelength 80
The absorbance at 8 ± 4 nm is x1, and the wavelength is 816 ±
The absorbance at 4 nm is x2, and the wavelength is 825 ± 4n
The absorbance at a wavelength of 833 ± 4 nm is x4, the absorbance at a wavelength of 841 ± 4 nm is x5, the absorbance at a wavelength of 850 ± 4 nm is x6, the absorbance at a wavelength of 858 ± 4 nm is x7, and the wavelength is 866 ± The absorbance at 4 nm is x
8, the absorbance at a wavelength of 875 ± 4 nm is x9, the absorbance at a wavelength of 883 ± 4 nm is x10,
The absorbance at a wavelength of 891 ± 4 nm is x11, the absorbance at a wavelength of 900 ± 4 nm is x12, and the wavelength is 9
The absorbance at 08 ± 4 nm is x13, and the wavelength is 917.
The absorbance at ± 4 nm is x14 and the wavelength is 925 ± 4
The absorbance at nm is x15 and the wavelength is 933 ± 4 nm
The absorbance at 942 ± 4 nm is x17, the absorbance at 950 ± 4 nm is x18, the absorbance at 958 ± 4 nm is x19, the absorbance at 967 ± 4 nm is x20, and the wavelength is 975 ± 4 nm. Absorbance at x21
And the absorbance at a wavelength of 983 ± 4 nm is x22, the absorbance at a wavelength of 992 ± 4 nm is x23,
The absorbance at a wavelength of 1000 ± 4 nm is x24, the absorbance at a wavelength of 1008 ± 4 nm is x25,
The absorbance at 17 ± 4 nm is x26, and the wavelength is 1025.
The absorbance at ± 4 nm is x27, and the wavelength is 1033 ± 4.
The absorbance at nm is x28, and the wavelength is 1042 ± 4 nm.
The absorbance at a wavelength of 1050 ± 4 nm is x30, the absorbance at a wavelength of 1058 ± 4 nm is x31, the absorbance at a wavelength of 1066 ± 4 nm is x32, and the absorbance at a wavelength of 1075 ± 4 nm is x33; By performing partial least squares regression analysis using x33 as an explanatory variable and the amylose content y (%) in the polished rice sample as an objective variable, the coefficient b1 is set to -222≤b1≤1790, and the coefficient b
2 is −544 ≦ b2 ≦ 2140, the coefficient b3 is −5100 ≦ b3 ≦ 1520, and the coefficient b4 is −103000 ≦ b4 ≦ 119.
0 and the coefficient b5 is -74700 ≦ b5 ≦ 35
500 and the coefficient b6 is -52700 ≦ b6 ≦
53700, and the coefficient b7 is −104000 ≦ b7 ≦
5550, and the coefficient b8 is -80600 ≦ b
8 ≦ 29100 and the coefficient b9 is −60700
≤ b9 ≤ 51200 and the coefficient b10 is -1070
00 ≦ b10 ≦ 5750, and coefficient b11 is −8
3700 ≦ b11 ≦ 33400, and the coefficient b12 is
-50600 ≦ b12 ≦ 58200, coefficient b13 is −1000000 ≦ b13 ≦ 6350, coefficient b
14 is −84200 ≦ b14 ≦ 24600, coefficient b15 is −60400 ≦ b15 ≦ 50700, and coefficient b16 is −98800 ≦ b16 ≦ 743.
0 and the coefficient b17 is −78700 ≦ b17 ≦ 33
600 and the coefficient b18 is −63900 ≦ b18 ≦
50200, and the coefficient b19 is -95400 ≦ b19 ≦
6070, and the coefficient b20 is −78000 ≦ b
20 ≦ 25600, and the coefficient b21 is −52300
≦ b21 ≦ 56000 and the coefficient b22 is −1030
00 ≦ b22 ≦ 2870, and the coefficient b23 is -7
8300 ≦ b23 ≦ 32400, and the coefficient b24 is
−62200 ≦ b24 ≦ 52300, coefficient b25 −99500 ≦ b25 ≦ 3690, coefficient b
26 is −77800 ≦ b26 ≦ 34600, the coefficient b27 is −59800 ≦ b27 ≦ 52300, and the coefficient b28 is −103000 ≦ b28 ≦ 409.
0 and the coefficient b29 is -80900 ≦ b29 ≦ 30
The coefficient b30 is set to −56000 ≦ b30 ≦
54300, and the coefficient b31 is −1690 ≦ b31 ≦
2980, and the coefficient b32 is −1710 ≦ b
32 ≦ 1540, and the coefficient b33 is −1980
≤ b33 ≤ 1400, and constant term b. To -2010
00 ≦ b. Assuming ≦ 1270000, the amylose content y (%) in the above-mentioned polished rice sample is expressed by the following equation: × x10 + b11 × x11 + b12 × x12 + b13 × x13 + b14 × x14 + b15 × x15 + b16 × x16 + b17 × x17 + b18 × x18 + b19 × x19 + b20 × x20 + b21 × x21 + b22 × x22 + b23 × x23 + b24 × x26 + b25 × x26 + b25 × x26 + b25 × x26 + b25 × x27 x32 + b33 x x33 + b. A device for measuring amylose content of rice, comprising a data analysis unit for calculating the content of rice. 6. The apparatus for measuring amylose content of rice according to claim 5, wherein the data analysis unit is configured to calculate the coefficient b1 = b2 = b3 = b31.
= B32 = b33 = 0, and the coefficient b4 is -2613.
3 ≦ b4 ≦ 957, and the coefficient b5 is -6
4320 ≦ b5 ≦ 25608, and the coefficient b6 is
−959 ≦ b6 ≦ 53699, and the coefficient b
7 is set to −21172 ≦ b7 ≦ 5550, the coefficient b8 is set to −71496 ≦ b8 ≦ 21526, and the coefficient b9 is set to −3954 ≦ b9 ≦ 511.
The coefficient b10 is set to -20859 ≦ b10 ≦
5748, and the coefficient b11 is -72463 ≦ b
11 ≦ 26610, and the coefficient b12 is
3 ≦ b12 ≦ 58193, and the coefficient b13 is −2.
0104 ≦ b13 ≦ 6350, and the coefficient b14 is
−73737 ≦ b14 ≦ 18320, and the coefficient b
15 is set to −4475 ≦ b15 ≦ 50700, the coefficient b16 is set to −18813 ≦ b16 ≦ 7428, and the coefficient b17 is set to −67999 ≦ b17 ≦ 251.
17 and the coefficient b18 is −6790 ≦ b18 ≦
50200, and the coefficient b19 is set to −22797 ≦ b
19 ≦ 5056, and the coefficient b20 is −7034
3 ≦ b20 ≦ 18703, and the coefficient b21 is
33 ≦ b21 ≦ 55996, and the coefficient b22 is
−23377 ≦ b22 ≦ 2866, and the coefficient b
23 is −68306 ≦ b23 ≦ 24348, the coefficient b24 is −5034 ≦ b24 ≦ 52266, and the coefficient b25 is −22473 ≦ b25 ≦ 36.
90, and the coefficient b26 is -68729 ≦ b26 ≦
26265, and the coefficient b27 is -3992 ≦ b
27 ≦ 52300, and the coefficient b28 is −2217
6 ≦ b28 ≦ 4090, and the coefficient b29 is −7
0068 ≦ b29 ≦ 21690, and the coefficient b30 is
0 ≦ b30 ≦ 54300, and the constant term b. −201000 ≦ b. An apparatus for measuring the amylose content of rice, wherein the value of ≦ 346450 is calculated and the amylose content y in the polished rice sample is calculated. 7. The apparatus for measuring amylose content of rice according to claim 6, wherein the data analysis unit is configured to calculate the coefficient b1 = b2 = b3 = b31.
= B32 = b33 = 0, and the coefficient b4 is -2713
≦ b4 ≦ −68, and the coefficient b5 is 50
1 ≦ b5 ≦ 5442, and the coefficient b6 is -9
59 ≦ b6 ≦ 2667, and the coefficient b7 is -4
059 ≦ b7 ≦ 5550, and the coefficient b8 is −
2586 ≦ b8 ≦ −658, and the coefficient b9 is
−3416 ≦ b9 ≦ −159, and the coefficient b10 is
−765 ≦ b10 ≦ 5158, the coefficient b11 is −2894 ≦ b11 ≦ 3586, and the coefficient b12 is
2199 ≦ b12 ≦ 4866, and the coefficient b
13 is −215 ≦ b13 ≦ 6350, the coefficient b14 is −4814 ≦ b14 ≦ 2213, the coefficient b15 is −4020 ≦ b15 ≦ −2287, the coefficient b16 is 2439 ≦ b16 ≦ 7428, and the coefficient b17 is 259 ≦ b17 ≦ 2271,
The coefficient b18 is set to −4561 ≦ b18 ≦ −1415, and the coefficient b19 is set to −1884 ≦ b19 ≦ 3974.
And the coefficient b20 is −1722 ≦ b20 ≦ −73
2 and the coefficient b21 is 1612 ≦ b21 ≦ 42
87 and the coefficient b22 is −2239 ≦ b22 ≦ 2
866, and the coefficient b23 is 665 ≦ b23 ≦
1840, and the coefficient b24 is −3436 ≦ b24 ≦
−884, and the coefficient b25 is −680 ≦ b25
≦ 3690, and the coefficient b26 is 463 ≦ b
26 ≦ 3438, and the coefficient b27 is −2176 ≦
b27 ≦ −445, and the coefficient b28 is −595
≤ b28 ≤ 4090, and the coefficient b29 is -212
1 ≦ b29 ≦ 768, and the coefficient b30 is
0 ≦ b30 ≦ 2004, and the constant term b. To -26
592 ≦ b. An apparatus for measuring amylose content of rice, wherein the amylose content y in the polished rice sample is calculated with ≦ 69. 8. The apparatus for measuring amylose content of rice according to claim 7, wherein the data analysis unit is configured to calculate the coefficient b1 = b2 = b3 = b31.
= B32 = b33 = 0, and the coefficient b4 is -2713≤b
4 ≦ −67, and the coefficient b5 is 666 ≦ b5
≦ 2473, and the coefficient b6 is −237 ≦ b6 ≦
1324, and the coefficient b7 is -518 ≦ b7 ≦
835, and the coefficient b8 is −1340 ≦ b8 ≦ −
658, and the coefficient b9 is -842≤b9≤-1.
59 and the coefficient b10 is −14 ≦ b10 ≦ 67
5 and the coefficient b11 is −1157 ≦ b11 ≦ 164.
The coefficient b12 is 2199 ≦ b12 ≦ 2624, the coefficient b13 is 750 ≦ b13 ≦ 2223, the coefficient b14 is −2890 ≦ b14 ≦ −2692, and the coefficient b15 is −3390 ≦ b15 ≦ −2287, The coefficient b16 is set to 2439 ≦ b16 ≦ 3340, the coefficient b17 is set to 1640 ≦ b17 ≦ 2088, the coefficient b18 is set to −2458 ≦ b18 ≦ -1415, the coefficient b19 is set to −1884 ≦ b19 ≦ -1406, and the coefficient b20 is −1533 ≦ b20 ≦ −732, the coefficient b21 is 3080 ≦ b21 ≦ 4287, the coefficient b22 is -1639 ≦ b22 ≦ 1120, the coefficient b23 is 665 ≦ b23 ≦ 1446, and the coefficient b24 is − 1120 ≦ b24 ≦ −884, the coefficient b25 is −680 ≦ b25 ≦ −87, and the coefficient b26 is 463 ≦ b26 ≦ 1279 The coefficient b27 is set to −1264 ≦ b27 ≦ −445, the coefficient b28 is set to −508 ≦ b28 ≦ 38, the coefficient b29 is set to 37 ≦ b29 ≦ 386, and the coefficient b30 is set to 311 ≦ b30 ≦ 938. Constant term b. −47 ≦ b. An apparatus for measuring amylose content of rice, wherein the amylose content y in the polished rice sample is calculated as ≦ −17.