JP2007187486A - Method and device for estimating component of material - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve reliability of an estimated value by enabling accurate estimation as much as possible, not only when the type of an imaging part is different but also when an imaging environment such as a temperature condition or a humidity condition is different. <P>SOLUTION: A specimen K having a prescribed quantity of material or an extract of the prescribed quantity of material is stored in a specimen container 1, and a plurality of reference specimens C wherein a known component quantity is different respectively, which are reference specimens C used as a reference for an equivalent configuration to the specimen K wherein a specific component quantity is known beforehand, are stored in a plurality of reference containers 2, respectively. The specimen K in the specimen container 1 and the reference specimen C in each reference container 2 are imaged simultaneously by the imaging part 20, and a mutual relation between the brightness of the reference specimen C and the component quantity of the material is calculated, and the brightness of the specimen K is detected based on an image imaged by the imaging part 20, and the quantity of the specific component among components in the material in the specimen K is calculated based on the brightness and the mutual relation. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a substance component estimation method for calculating the amount of various components contained in a substance from color elements related to the substance in various other substances including agriculture, such as soil and crops grown on the soil. And a component estimation apparatus for substances.

  Conventionally, as a component estimation device for this kind of substance, there is an estimation device that the applicant of the present application previously proposed and calculates the amount of component contained in the soil from the color element of the soil in the soil as the material (Patent Document 1). ).

  As shown in FIG. 14, this is to estimate the total carbon amount of carbon and the total nitrogen amount of nitrogen among the components contained in the soil, which is a substance. A transparent container in which a predetermined amount of soil is arranged in a plurality of rows 100. The container 100 is provided integrally with a white tray 101 that holds the bottom wall exposed. A color bar 102 serving as a reference for luminance is attached to the back surface (the CCD sensor side described later) of the tray 101. This tray 101 is placed on a base 104 of a scanner 103 as an imaging unit. The scanner 103 irradiates the bottom of the container 100 on the base 104 and the tray 101 with light from a fluorescent tube or LED from below the base 104, and reads the RGB brightness by a CCD sensor or the like (not shown) to reflect the reflected light. And is captured as a digital image.

Then, the scanner 103 is operated to image the container 100 and the tray 101 on the base 104, that is, the image is acquired by scanning the surface of the soil and the color bar 102 through the bottom of the container 100. Is output to the calculation unit 105 having the above function. In the calculation unit 105, first, an image of only the soil surface of R luminance is extracted for each container 100. On the other hand, the color bar 102 image is extracted, and white, gray, and black R luminances are calculated from the color bar 102 image. Next, the R luminance of each pixel of the image on the soil surface is detected, and the detected R luminance of each pixel is corrected with reference to the R luminance of the color bar, and R constituting each pixel of the corrected image is formed. All the luminance values are added together, and the average R luminance is calculated by dividing by the total number of pixels constituting this image.
In this case, since the extracted image is corrected based on the R luminance of the color bar 102, for example, the performance of the CCD or the like varies depending on the manufacturer, type, and type of the scanner 103, and is acquired from the scanner 103. In the obtained image, even if there is a difference in luminance of each pixel, this can be corrected, and the detection accuracy can be improved.

Next, in the calculation unit 105, the correlation between the detected average R luminance of the soil surface for each container 100, the luminance of the soil surface serving as a reference, and the amount of components in the soil serving as the reference stored in the storage unit in advance. From the relationship, the total amount of carbon and the total amount of nitrogen as component amounts in the soil for each container 100 are calculated.
Thus, the component amount is calculated simply by putting the soil in the container 100 and taking an image with the scanner 103. Therefore, almost the same component as the component amount in the soil can be obtained by anyone without experience or skill. The amount can be estimated.

JP 2005-17115 A

  By the way, in the above-described conventional component estimation apparatus, the color bar 102 is simultaneously imaged together with the soil as the specimen, and the extracted image is corrected based on the luminance of the color bar 102. This is effective when the model is different, but the reliability of the estimated value may not always be sufficiently secured when the imaging environment is different, for example, when the specimen is imaged on different days. was there. For example, when the temperature condition and the humidity condition are different as the imaging environment, the color of the color bar 102 does not change, but the color of the specimen may change, resulting in an error. Further, the color itself of the color bar 102 may change over time, which also causes a correction error.

  The present invention has been made in view of the above-described problems, and it is possible to estimate as accurately as possible even in a case where the imaging environment such as a temperature condition and a humidity condition is different as well as in a case where the model of the imaging unit is different. Thus, an object of the present invention is to provide a substance component estimation method and substance component estimation apparatus that improve the reliability of an estimated value.

In order to achieve such an object, the present invention accommodates a specimen having a predetermined amount of substance or an extract of a predetermined amount of substance in a specimen container, and images the surface of the specimen contained in the specimen container. The component of the substance that detects the luminance of the surface of the specimen based on the image captured by the imaging unit and calculates the component amount of the specific component among the contained components in the substance based on the detected luminance In the estimation method,
A plurality of reference specimens, each of which is a reference specimen in a form equivalent to the above-described specimen whose amount of a specific component is known in advance, each having a different amount of the known component, are housed in a plurality of reference containers, respectively. The imaging unit simultaneously images the surface of the sample accommodated in the sample container and the surface of the reference sample accommodated in each reference container, and each reference reference is based on the image captured by the imaging unit. Detecting the luminance of the surface of the reference sample contained in the container, calculating a correlation between the luminance of the surface of the reference sample and the amount of a specific component related to the reference sample based on the detected luminance, and performing the imaging The luminance of the surface of the sample contained in the sample container is detected based on the image captured by the unit, and the component amount of the specific component related to the sample is determined from the detected luminance of the surface of the sample and the correlation. It is configured to calculate.

  Thus, the imaging unit simultaneously images the surface of the sample and the surface of the reference sample, detects the luminance of the surface of the reference sample for each imaging, and the luminance of the surface of the reference sample and the component of the substance related to the reference sample Since the correlation with the component amount is calculated, not only when the model of the imaging unit is different, but also when the imaging environment such as the temperature condition and the humidity condition is different, the model condition and the imaging environment each time. The correlation according to the change can be calculated, so that the component amount of the component calculated from the luminance based on the image of the surface of the specimen imaged by the imaging unit and the above correlation is the model condition or imaging environment change Thus, an accurate estimated value corresponding to is obtained, and the reliability of the estimated value is improved.

In this case, the sample is configured to include a predetermined amount of the substance, and the reference sample is configured to include the same type of substance as the predetermined amount of the sample in which the component amount of the specific component is known in advance. can do.
In addition, the specimen may be configured with an extract of a predetermined amount of the substance, and the reference specimen may be configured with a reagent in which the amount of the specific component is known in advance by mixing the specific component. .

  In addition, if necessary, the substance is a plant. Moreover, it is set as the structure whose said substance is soil as needed. The amount of components of various substances related to agriculture such as soil and crops growing on the soil can be estimated, and in particular, it can be used in the agricultural field.

Further, in order to achieve the above object, the present invention images a specimen container containing a specimen having a predetermined amount of substance or a predetermined amount of substance extract, and a surface of the specimen contained in the specimen container. An imaging unit that detects the luminance of the surface of the specimen based on the image captured by the imaging unit and calculates a component amount of a specific component among the components contained in the substance based on the detected luminance In the substance component estimation apparatus comprising:
A plurality of reference containers each containing a plurality of reference samples, each of which is a reference sample in a form equivalent to the above-described sample whose component amount of a specific component is known, and each of which has a different known component amount The imaging unit is configured to simultaneously image the surface of the specimen housed in the specimen container and the surface of the reference specimen contained in each of the reference containers;
Based on the luminance detected by the reference luminance detecting means, the reference luminance detecting means for detecting the luminance of the surface of the reference specimen contained in each reference container based on the image picked up by the imaging section, Correlation calculating means for calculating the correlation between the luminance of the surface of the reference sample and the component amount of the specific component related to the reference sample, and the sample container accommodated in the sample container based on the image captured by the imaging unit A sample luminance detecting unit for detecting the luminance of the surface of the sample, a luminance of the surface of the sample detected by the sample luminance detecting unit, and a correlation calculated by the correlation calculating unit; And a component amount calculating means for calculating the component amount.

  As a result, the imaging unit simultaneously images the surface of the specimen and the surface of the reference specimen, detects the brightness of the surface of the reference specimen for each imaging, and determines the brightness of the surface of the reference specimen and a specific component related to the reference specimen. Since the correlation with the component amount is calculated, not only when the model of the imaging unit is different, but also when the imaging environment such as the temperature condition and the humidity condition is different, the model condition and the imaging environment each time. The correlation according to the change can be calculated. Therefore, the component amount of the specific component calculated from the luminance based on the image of the surface of the specimen imaged by the imaging unit and the above correlation is the model condition or imaging It becomes an accurate estimated value according to the environmental change, and the reliability of the estimated value is improved.

  And the said imaging part is set as the structure which images multiple said sample containers as needed. Since image processing is performed on a plurality of specimens, and the amount of a specific component for each specimen can be calculated at a time, the processing efficiency is improved.

In addition, if necessary, the specimen container and the reference container formed of a transparent material are used, and the imaging unit includes a base on which the specimen container and the reference container are placed, and the base A scanner that scans the sample container and the reference container and acquires an image through the bottom wall of each container;
The calculation unit includes reference image extraction means for extracting an image of a predetermined range of an image of only the surface of the reference specimen of the reference container from the image output from the scanner, and the image output from the scanner A sample image extracting means for extracting an image of a predetermined range that is an image of only the surface of the sample in the sample container,
The reference luminance detecting means of the calculating unit includes a pixel luminance detecting means for detecting the luminance of each pixel in the image extracted from the reference image extracting means, and an average from the luminance of each pixel detected by the pixel luminance detecting means. An average luminance calculating means for calculating luminance,
The sample luminance detection unit of the calculation unit includes a pixel luminance detection unit that detects the luminance of each pixel in the image extracted from the sample image extraction unit, and an average from the luminance of each pixel detected by the pixel luminance detection unit. An average luminance calculating means for calculating luminance,
The correlation calculation means is configured to calculate the correlation based on the average brightness for the reference specimen of each reference container calculated by the average brightness calculation means of the reference brightness detection means,
The component amount calculation unit is configured to calculate the component amount based on the average luminance of the sample in the sample container calculated by the average luminance calculation unit of the sample luminance detection unit.

  The image extracting means extracts an image of only the specimen surface for each container and extracts an image within a predetermined range, and obtains the average luminance of each pixel of the image, thereby reducing errors and preventing variations in detected luminance. . In addition, the scanner may be a commercially available one, and an existing one can be used, so that the introduction cost can be reduced.

  Further, if necessary, a plurality of each of the above containers is held with the bottom wall exposed and can be placed on the base of the imaging unit, and the scanner scans the bottom walls of the plurality of containers provided on the tray. It is configured to do. Since a plurality of containers are held by a tray, handling such as transportation and imaging becomes easy. In addition, the tray can block light from the outside, and more accurate imaging can be performed.

In addition, if necessary, the sample is configured to have a predetermined amount of substance, and the reference sample has the same kind of substance as the predetermined amount of the sample whose amount of a specific component is known in advance. Can be configured.
In addition, the specimen may be configured with an extract of a predetermined amount of the substance, and the reference specimen may be configured with a reagent in which the amount of the specific component is known in advance by mixing the specific component. .

  In addition, if necessary, the substance is a plant. Moreover, it is set as the structure whose said substance is soil as needed. The amount of components of various substances related to agriculture such as soil and crops growing on the soil can be estimated, and in particular, it can be used in the agricultural field.

According to the substance component estimation method and substance component estimation apparatus of the present invention, the imaging unit simultaneously images the surface of the specimen and the surface of the reference specimen, and detects the luminance of the surface of the reference specimen for each imaging. Since the correlation between the brightness of the surface of the reference sample and the amount of the component of the substance related to the reference sample is calculated, the imaging environment such as the temperature condition and humidity condition can be used as well as when the model of the imaging unit is different. Even if they are different, it is possible to calculate the correlation according to these model conditions and imaging environment changes each time. Therefore, from the luminance based on the image of the surface of the specimen imaged by the imaging unit and the above correlation The calculated component amount becomes an accurate estimated value corresponding to the model condition and the imaging environment change, and the reliability of the estimated value can be improved.
As a result, since the estimation of this component amount is not based on the subjective experience that humans visually observe, it can be estimated almost uniquely without causing individual differences, and it is stable without causing variations in the estimated amount. It can be expected to be used in the agricultural field.

  Hereinafter, a substance component estimation method and a substance component estimation apparatus according to embodiments of the present invention will be described in detail with reference to the accompanying drawings. Since the substance component estimation method according to the embodiment of the present invention is realized by the substance component estimation apparatus according to the embodiment of the present invention, in the operation of the substance component estimation apparatus according to the embodiment of the present invention. explain.

As shown in FIG. 1 to FIG. 3, the substance component estimation apparatus according to the embodiment of the present invention includes a specimen container 1 that contains a specimen K having a predetermined amount of substance or a predetermined amount of substance extract. I have. A plurality of specimen containers 1 are prepared.
Further, a plurality of reference samples C each having a reference component C in a form equivalent to the sample K in which the component amount of a specific component is known in advance, each containing a plurality of reference samples C having different known component amounts. The reference container 2 is provided.

  As shown in FIG. 2, the sample container 1 and the reference container 2 are transparent containers of the same material and shape, and are formed into a cup shape with a transparent material such as glass, plastic resin, or quartz glass. Yes. For example, a container having an outer diameter of 32 mm, an inner diameter of 27 mm, a height of 15 mm, and a wall thickness of 1.3 mm is used.

The containers 1 and 2 are held on the tray 10. As shown in FIGS. 1 and 3, the tray 10 is formed on a base plate 21 of a scanner 23, which will be described later, and is formed into a resin-made rectangular plate with a size covering the container. It has holes 11 in the form of a matrix to be held (in the embodiment, 10 in 2 rows and 5 columns). The tray 10 is colored gray.
In each hole 11 of the tray 10, four reference containers 2 and six sample containers 1 are held. In the tray 10, the numbers of the containers 1 and 2 are attached to the front side of the tray 10 in the vicinity of the holes 11 so that the sample K placed in the containers 1 and 2 can be easily identified. 1 to 6 are assigned as the numbers of the sample containers 1, and S1 to S4 are assigned as the numbers of the reference containers 2.

Here, the sample K and the reference sample C are desirably in the form of powder, liquid, or a fluid mixture of powder and liquid.
In addition, the specimen K can be configured to have a predetermined amount of substance. In this case, the reference sample C can be configured by having a predetermined amount of the same kind of substance as the sample K in which the component amount of the specific component is known in advance. For example, the stems and leaves of paddy rice described later (referred to as stems and leaves on the ground among strains arbitrarily selected from a large number of paddy rice plants planted in a predetermined paddy field. In some cases, the ears are removed.) In some cases, the mass percentage of nitrogen, which is the amount of components in the foliage, may be obtained. As a substance, only rice leaves or stems can be selected.
In addition, the specimen K can be configured with an extract of a predetermined amount of substance. In this case, the reference sample C can be composed of a reagent in which a specific component is mixed and the amount of the specific component is known in advance. For example, when the substance is soil described later, the available phosphate content (mg) per 100 g of soil may be obtained. In this case, the amount of components can be estimated by coloring phosphorus in the soil extraction solution by a chemical reaction and analyzing luminance information of the color developing solution. In addition to phosphoric acid, the amount of each component can be estimated by coloring the nitrogen, potassium, calcium, and magnesium contents in the soil extraction solution by a chemical reaction and analyzing the luminance information of the coloring solution.

The substance component estimation apparatus according to the embodiment of the present invention simultaneously captures the surface of the specimen K accommodated in the specimen container 1 and the surface of the reference specimen C accommodated in the reference container 2. It has.
Specifically, the imaging unit 20 includes a base 21 formed of a glass plate on which the tray 10 holding the containers 1 and 2 is placed in contact with the bottom walls 3 of the containers 1 and 2, and a tray on the base 21. 10 and a cover 22 that covers the opening side of the containers 1 and 2 and that can be opened and closed with respect to the base 21, and a scanner 23 that scans the tray 10 on the base 21 to capture an image. Yes. The base 21 is formed of, for example, an A4 size document placing table.

The scanner 23 irradiates the tray 10 on the base 21 and the bottom wall 3 of the containers 1 and 2 with the light of the fluorescent tube and the LED from below the base 21, reads the RGB luminance by the CCD sensor or the like from the reflected light, The RGB brightness of the reference specimen C and specimen K is acquired through the bottom walls 3 of the containers 1 and 2 and captured as digital images. R luminance, G luminance, and B luminance are each 256 steps from 0 to 255, and the pixels of the image are composed of combinations of the RGB luminances, and about 16 million colors are expressed.
The scanner 23 may be a commercially available one, and since an existing one is used, the introduction cost can be kept low.

  Furthermore, the substance component estimation apparatus according to the embodiment of the present invention includes a calculation unit 30 that operates by a function of a CPU or the like. The calculation unit 30 detects the luminance of the surface of the specimen K based on the image captured by the imaging unit 20, and calculates the component amount of a specific component to be estimated for the contained component in the substance based on the detected luminance. To do.

  Specifically, the calculation unit 30 includes a reference image extraction unit 31, a sample image extraction unit 32, a reference luminance detection unit 33, a correlation calculation unit 34, a sample luminance detection unit 35, and a component amount calculation unit 36. It is prepared for.

The reference image extracting unit 31 extracts an image of only the surface of the reference sample C of each reference container 2 from the image output from the scanner 23 and having a predetermined range e. For example, as shown in FIG. 2A, a square having a center of gravity at the center of the bottom wall 3 surface of the containers 1 and 2 (for example, a square having a side of 15 mm) is defined as a predetermined range e (analysis region).
The sample image extraction unit 32 extracts an image of only the surface of the sample K of the sample container 1 from the image output from the scanner 23 and an image in a predetermined range e. For example, as described above, as shown in FIG. 2A, a square having a center of gravity at the center of the bottom wall 3 surface of the containers 1 and 2 (for example, a square having a side of 15 mm) is defined as a predetermined range e (analysis region). To do.
More specifically, each of the image extraction means 31 and 32 is based on the difference in luminance between the back surface of the tray 10 and the surface of the specimen contained in the reference container 2 and the specimen container 1 and the shape of the specimen container. The images of only the surfaces of the containers 1 and 2 are extracted in each case, and an image in a predetermined range is extracted from the surface images of the containers 1 and 2.

  The reference luminance detection unit 33 detects the luminance of the surface of the reference sample C accommodated in each reference container 2 imaged by the imaging unit 20. The reference luminance detecting unit 33 includes a pixel luminance detecting unit 33a for detecting the luminance of each pixel in the image extracted from the reference image extracting unit 31, and an average luminance from the luminance of each pixel detected by the pixel luminance detecting unit 33a. And an average luminance calculating means 33b for calculating. The pixel luminance detecting unit 33a detects the R, G, and B luminances of each pixel in an image in a predetermined range on the surface of the reference specimen C extracted from the reference image extracting unit 31. This value is 256 steps from 0 to 255. The average luminance calculating means 33b adds the luminance for each of R, G, and B of each pixel of the image in the predetermined range, and divides by the number of pixels constituting the image in the predetermined range, and averages for each R, G, and B The brightness is calculated.

  The correlation calculation unit 34 calculates the correlation between the luminance of the surface of the reference sample C and the component amount to be estimated of the substance related to the reference sample C based on the luminance detected by the reference luminance detection unit 33. When the plurality of reference samples C having different specific component amounts selected above are scanned by the scanner 23, the reference image extraction unit 31 extracts a predetermined range of images, and the reference luminance detection unit 33 averages the correlation. Luminance is calculated, and a correlation between the measured value and the calculated average luminance is determined by, for example, a prediction formula using the least square method.

  Similar to the reference luminance detection unit 33, the sample luminance detection unit 35 detects the luminance of the surface of the sample K stored in each sample container 1 imaged by the imaging unit 20. The sample luminance detecting unit 35 includes a pixel luminance detecting unit 35a for detecting the luminance of each pixel in the image extracted from the sample image extracting unit 32, and an average luminance from the luminance of each pixel detected by the pixel luminance detecting unit 35a. And an average luminance calculating means 35b for calculating. The pixel luminance detecting unit 35a detects R, G, and B luminances of each pixel in an image in a predetermined range on the surface of the sample K extracted from the sample image extracting unit 32. This value is 256 steps from 0 to 255. The average luminance calculating means 35b adds the luminance for each of R, G, and B of each pixel of the image in the predetermined range, and divides by the number of pixels constituting the image in the predetermined range, and averages for each of R, G, and B The brightness is calculated.

The component amount calculation unit 36 calculates the component amount of a specific component related to the sample K from the surface luminance of the sample K detected by the sample luminance detection unit 35 and the correlation calculated by the correlation calculation unit 34. To do.
Then, the output device 37 such as a display monitor or a printer connected to the calculation unit 30 displays a table in which the number of each container is written together with the luminance and the component amount.

Therefore, when estimating the amount of a component using the component estimation apparatus of this substance, it is performed as follows.
First, a case where the specimen K is configured to have a predetermined amount of substance will be described. In this case, the reference sample C is configured to have a predetermined amount of the same kind of substance as the sample K in which the component amount of a specific component is known in advance.
For example, when the substance is paddy rice and the mass percentage of nitrogen, which is the component amount of the paddy rice foliage, is determined, first the rice foliage is collected. The stem and leaf part refers to a part of a stem and a leaf that is on the ground, among strains arbitrarily selected from a large number of strains of paddy rice planted in a paddy field of a predetermined section. If there are spikes, remove the spikes. In addition, a plurality of strains (for example, 3 strains) are collected from a paddy field in a predetermined section for averaging. And as the specimen K, using all of the foliage parts of these multiple strains, it is dried for a predetermined time with a ventilator to make the water content approximately 0% by mass, the dried foliage part is pulverized with a pulverizer, and a predetermined amount is obtained. It is assumed that a predetermined amount of ethanol is added to the specimen container 1.
Further, in the reference sample C, a sample of the foliage is collected from a paddy field having a different growth state by the same method as described above, and the mass percentage of nitrogen as the component amount of this sample is determined by a well-known quantitative analysis method (for example, Then, after wet ashing, the nitrogen content is determined by the indophenol method) and is accurately obtained in advance as an actually measured value. Then, a plurality of reference samples C having different components in stages are selected from various samples. Similarly, a predetermined amount is placed in the reference container 2 and a predetermined amount of ethanol is added.

  The reason for adding alcohol is to emphasize the color difference of the powder sample in order to prevent irregular reflection due to unevenness of the specimen surface.

Next, the four reference containers 2 of the reference sample C are placed on the tray 10, and the six sample containers 1 of appropriate specimens K such as those of the same lot or different lots are placed as the specimen K. . Then, the tray 10 is placed on the base 21 of the scanner 23 and covered with the cover 22. When carrying the container, the handleability can be improved by holding the tray 10.
Then, the scanner 23 is operated to take an image of the surface of the tray 10, and this image is output to the calculation unit 30.

Then, the calculation unit 30 is operated to obtain an estimated amount of average R luminance and nitrogen amount for each specimen container.
Specifically, when the calculation unit 30 operates, first, the reference image extraction unit 31 obtains an image of only the surface of the reference specimen C of each reference container 2 from the image output from the scanner 23 and an image in a predetermined range e. Extract.
The reference luminance detection unit 33 detects the luminance of the surface of the reference sample C accommodated in each reference container 2 imaged by the imaging unit 20. In the reference luminance detection means 33, the pixel luminance detection means 33a detects the R, G, B luminance of each pixel in the image in a predetermined range on the surface of the reference specimen C extracted from the reference image extraction means 31. The average luminance calculation means 33b adds the luminances of R, G, and B of each pixel of the image in the predetermined range, and divides by the number of pixels constituting the image of the predetermined range e for each of R, G, and B. The average brightness is calculated. In the embodiment, the average R brightness for each reference container 2 is obtained for nitrogen.

  Then, the correlation calculation unit 34 calculates the correlation between the luminance of the surface of the reference sample C and the component amount to be estimated of the substance related to the reference sample C based on the luminance detected by the reference luminance detection unit 33. When the plurality of reference specimens C having different selected specific component amounts are scanned by the scanner 23, the reference image extraction unit 31 extracts an image in a predetermined range e, and the reference luminance detection unit 33 The average luminance is calculated, and the correlation between the actually measured value and the calculated average luminance is determined by, for example, a prediction formula using the least square method.

  On the other hand, the sample image extraction means 32 extracts an image of only the surface of the sample K of the sample container 1 from the image output from the scanner 23 and an image in a predetermined range e. Similar to the reference luminance detection unit 33, the sample luminance detection unit 35 detects the luminance of the surface of the sample K stored in each sample container 1 imaged by the imaging unit 20. In the sample luminance detecting means 35, the pixel luminance detecting means 35a detects the R, G, B luminance of each pixel in the image in the predetermined range on the surface of the sample K extracted from the sample image extracting means 32. . Further, the average luminance calculating means 35b adds the luminance for each R, G, B of each pixel 24 of the image in the predetermined range, and divides by the number of pixels constituting the image in the predetermined range e to obtain R, G, The average brightness is calculated for each B. In the embodiment, the average R luminance for each specimen container 1 is obtained for the nitrogen amount.

Then, the component amount calculation unit 36 determines the component amount of a specific component related to the sample K from the surface luminance of the sample K detected by the sample luminance detection unit 35 and the correlation calculated by the correlation calculation unit 34. Is calculated.
Then, on the output device 37 such as a display monitor or a printer connected to the calculation unit 30, a table in which the luminance and the component amount are written together with the numbers of the containers 1 and 2 is displayed.

  In this case, the imaging unit 20 images the surface of the sample K and the surface of the reference sample C at the same time, detects the luminance of the surface of the reference sample C for each imaging, and the luminance of the surface of the reference sample C and the reference sample Since the correlation with the component amount of the specific component related to C is calculated, not only when the model of the imaging unit 20 is different, but also when the imaging environment such as the temperature condition and the humidity condition is different, each time, Correlation according to these model conditions and imaging environment changes can be calculated. Therefore, a specific component calculated from the luminance based on the image of the surface of the specimen K captured by the imaging unit 20 and the above correlation This component amount becomes an accurate estimated value according to the model conditions and the imaging environment change, and the reliability of the estimated value is improved.

  Further, the component amount of the specific component is calculated simply by putting the sample K in the container 1 and taking an image with the scanner 23, so that the component of the specific component can be performed by anyone without experience or skill. The amount can be estimated. In addition, the estimation of the amount of the component is not based on the subjective experience that is visually observed by humans. Therefore, there is no individual difference in the measurement of the amount of the component, and the amount of the component may vary. It is stable and reliability is improved. Furthermore, since the images of the plurality of sample containers 1 acquired by the scanner 23 can be processed at a time, the processing efficiency can be improved.

  Moreover, since the RGB luminance is calculated from the R luminance, the nitrogen amount of the rice can be stably estimated as an optimum parameter. The nitrogen content estimated in this way can be used as a reference during the cultivation management such as topdressing.

Next, a case where the specimen K is configured to have a predetermined amount of substance extract will be described. In this case, the reference sample C is composed of a reagent in which a specific component is mixed to make the amount of the specific component known in advance.
For example, when the substance is soil, and when the available phosphate content (mg) of the soil is determined as the content (mg) per 100 g of dry soil, the soil is collected and dried for a required time with an air dryer. The water content is reduced to about 0% by mass, the dried soil is pulverized with a mortar, the particle size is made uniform with a round-hole sieve, a predetermined amount of soil passed through the sieve is taken into an Erlenmeyer flask, and a predetermined amount of sulfuric acid is added to the predetermined amount. In addition, after shaking, filter. A predetermined amount of the filtrate is placed in a volumetric flask, a predetermined amount of demineralized water is added, a predetermined amount of the mixed color solution is added, and a predetermined amount of the above-mentioned demineralized water is further added and mixed well. A predetermined amount of the prepared mixed solution is put into a container and this is used as a specimen K.
On the other hand, the reference sample C is prepared by diluting a phosphorus standard solution. The reference sample C to be prepared is prepared, for example, with a phosphorus content of four stages, a predetermined amount of mixed color developing solution is added, and a predetermined amount is put in a container.

Then, measurement is performed by the same operation as described above. The component amount calculation unit 36 of the calculation unit 30 uses the luminance of the surface of the sample K detected by the sample luminance detection unit 35 and the correlation calculated by the correlation calculation unit 34 to determine the specific component related to the sample K. The component amount is calculated.
Then, on the output device such as a display monitor or printer connected to the calculation unit 30, a table in which the luminance and the component amount are written together with the number of each container is displayed.
In addition, since the sample K is prepared by taking out a predetermined amount from a predetermined amount of soil extract, the available phosphate content in the soil (content per 100 g of dry soil) is a measured value. Perform the required conversion for. The conversion is performed by the component amount calculation means 36 of the calculation unit 30.
As described above, even when the specimen K is configured to include an extract of a predetermined amount of substance, the same operations and effects as described above can be obtained.

Next, examples will be described.
[Example 1] (The nitrogen content is estimated from the brightness information of the dry powder of the foliage of paddy rice)
First, prepare rice samples. Rice harvesting conditions were as follows.
Sampling time: From mid-June to just before harvesting Sampling and harvesting method: Three rice plants (those that are planted continuously) are harvested from a paddy field in one predetermined section. Collect from the stock source. After mid-August, the ears were removed and analyzed.
Moreover, about the sample of S1-S4, the foliage part was extract | collected from the paddy field of the predetermined division from which a growth state differs similarly to said method. The paddy field from which the sample was collected was selected appropriately from the length of the rice plant, the color of the leaves, and the like.
Then, the collected base mud is thoroughly washed and dried with a draft dryer. After drying, 3 strains are pulverized with a pulverizer. Crush the 3 stocks together.
More specifically, drying is performed at 70 ° C. for 72 hours with a ventilating dryer (DS400 manufactured by Yamato Scientific Co., Ltd.). The dried rice is pulverized to a pulverized particle size of 1.0 mm by a wheely pulverizer (1029-C, manufactured by Yoshida Seisakusho). The pulverized product is put into a small petri dish (FS-30 manufactured by Flat) as a container. The petri dish has an outer diameter of 32 mm, an inner diameter of 27 mm, a height of 15 mm, and a wall thickness of 1.3 mm. The thickness of the sample is about 5 mm and the weight is 2 g. 2 ml of ethanol (manufactured by Wako Pure Chemical Industries) is added to the sample placed in the petri dish.
Place the petri dish on the special tray 10 (A4 size made of urethane) and place it on the scanner 23. A petri dish containing a reference sample C with a known nitrogen content is fitted into the holes S1 to S4 of the tray 10. A petri dish containing a sample K whose nitrogen content is unknown is inserted into holes 1 to 6 of the tray 10.

Then, an analysis program for the apparatus is activated to obtain an image in which a petri dish is fitted in the tray 10. The resolution at the time of imaging is 150 dpi. The calculation unit 30 analyzes the color of the sample in each petri dish among the obtained images.
The calculation unit 30 automatically recognizes 10 petri dishes from the difference in luminance between the petri dish and the tray 10. For each of the 10 petri dishes, a square having a center of gravity at the center of the circle on the bottom of the petri dish and having a side of 15 mm is set as the analysis region. For each pixel in the region to be analyzed, the luminance value of the RGB component is calculated. The average value of the RGB values of each pixel in the region to be analyzed is set as the R value, G value, and B value of the sample. A calibration curve is created using the reference sample C4 with a known nitrogen content, and the nitrogen content of the sample K6 with an unknown nitrogen content is estimated using the calibration curve.

In Example 1, two different types of nitrogen content (mass%) were selected from “Akitakomachi” and “Hitomebore” as reference samples C (S1 to S4). As the unknown specimen K, 12 samples (1-12) of “Akitakomachi” (produced in 2005) and 12 samples (13-24) of “Hitomebore” (produced in 2005) were used. In order to see the correlation, the nitrogen content of the unknown sample K was also measured. FIG. 4 shows a graph in which the measurement data I (August data) of Example 1 and the estimated numerical value (estimated value 1) are plotted on the calibration curve.
From this result, extremely high correlation is recognized, and it is recognized that the accuracy of the estimated value is high.

Next, in Example 1, using the same sample as above, the same measurement was performed with different days (October after 2 months). FIG. 5 shows a graph in which the measurement data II (October data) and the estimated value (estimated value 2) are plotted on the calibration curve.
From this result, it is recognized that an extremely high correlation is recognized and the accuracy of the estimated value is high.

FIG. 6 shows a numerical value (estimated value 3) obtained by estimating the nitrogen content from the R luminance value of the unknown sample K of the measured data II (October data) using the result of the measured data I as a calibration curve.
And the correlation was compared about said estimated value 1,2,3. 7 to 9 show correlation diagrams for the estimated values 1, 2, and 3.
From these results, the estimated values 1 and 2 have a high correlation because the sample K and the reference sample C are imaged at the same time, and the component amount is estimated by the calibration curve created by detecting the luminance of the reference sample C for each imaging. Compared to those seen (FIGS. 7 and 8), the estimated value 3 uses a different calibration curve, so the correlation of the estimated value 3 is inferior to these (FIG. 9). From this, compared to determining the calibration curve uniquely, it is of course possible to create a calibration curve and estimate the component amount for each measurement, when the model of the imaging unit 20 is different, Even when the imaging environment such as temperature and humidity conditions are different, the estimation can be made according to these model conditions and changes in the imaging environment to obtain a more accurate estimated value, and the reliability of the estimated value is improved. It can be seen that it is excellent.

[Example 2] (Estimation of available phosphate content of soil)
(1) Preparation of specimen K Soil is collected and dried at 70 ° C. for 72 hours with a ventilator (DS400 manufactured by Yamato Scientific Co., Ltd.). The dried soil is pulverized with a mortar, and the particle size is made uniform with a circular hole sieve having a mesh size of 2 mm. 1 g of the soil passed through the sieve is taken and put into a 300 ml Erlenmeyer flask. Add 200 ml of 0.002N sulfuric acid (pH 3.0) and shake for 30 minutes. After shaking, Toyo Filter Paper No. Filter through 5B. Take 20 ml of the filtrate into a 50 ml volumetric flask and add 20 ml of demineralized water. To this, 8 ml of the mixed color developer is added. The mixed color solution consists of 5N sulfuric acid, 4% ammonium molybdate, 0.1M ascorbic acid solution, and ammonium potassium tartrate solution. Add demineralized water to the mark of the volumetric flask and mix well. 2 ml of the prepared specimen K is placed in the same petri dish as described above.

(2) Preparation of reference sample C On the other hand, a standard sample for preparing a calibration curve is prepared by separately diluting a phosphor standard solution. The phosphorus content of the prepared standard sample is 0 ppm, 0.2 ppm, 0.4 ppm, and 0.8 ppm. Take 20 ml of each of these in a 50 ml volumetric flask and add 20 ml of demineralized water. To this, 8 ml of the mixed color developer is added. Add demineralized water to the mark of the volumetric flask and mix well. 2 ml of the prepared reference sample C is placed in the above petri dish.

  Place the petri dish on the special tray 10 (A4 size made of urethane) and place it on the scanner 23. A petri dish containing a reference sample C with a known phosphorus content is fitted into the holes S1 to S4 of the tray 10. A petri dish containing a sample K with an unknown phosphorus content is inserted into the first through sixth holes of the tray 10.

Then, an analysis program for the apparatus is activated to obtain an image in which a petri dish is fitted in the tray 10. The resolution at the time of imaging is 150 dpi. The calculation unit 30 analyzes the color of the sample in each petri dish among the obtained images.
The calculation unit 30 automatically recognizes 10 petri dishes from the difference in luminance between the petri dish and the tray 10. For each of the ten petri dishes, a square having a center of gravity at the center of the circle on the bottom of the petri dish and having a side of 15 mm is set as the analysis region. For each pixel in the region to be analyzed, the luminance value of the RGB component is calculated. The average value of the RGB values of each pixel in the region to be analyzed is set as the R value, G value, and B value of the sample. A calibration curve is created from the reference sample C4 having a known phosphorus content, and the phosphorus content of the sample K6 having an unknown phosphorus content is estimated using the calibration curve.

In Example 2, the soil extract of the test field in Iwate Agricultural Research Center, which was investigated in 2005 as the unknown specimen K, was used. Twenty-four unknown specimens K were measured as described above.
FIG. 10 shows measurement data of the reference sample C of Example 2. FIG. 11 shows a correlation diagram between actual measurement values and estimated numerical values for 24 unknown specimens K. From this result, extremely high correlation is recognized, and it is recognized that the accuracy of the estimated value is high.
The analytical value represents how much (in ppm) phosphorus is contained in the soil extract using sulfuric acid. This value (ppm) is converted into a value indicating how many mg of phosphoric acid is contained per 100 g of dry soil.

<Experimental example>
Next, the luminance of the same reference sample C was measured at different times or days. In FIG. 12, with respect to the reference specimen C of Example 1 above, the measured luminance value immediately after the start of the apparatus (FIG. 12A) and the measured luminance value for the 20th time after starting (FIG. 12B) The results are shown. From this result, it can be seen that even with the same component content, the luminance is slightly different for each measurement, and it is effective to estimate the component amount by creating a calibration curve for each measurement.
In addition, FIG. 13 shows the first measured luminance value (FIG. 13A) for the reference sample C of Example 2 and the measured luminance value (FIG. 13 (FIG. 13 (A)) with different measurement dates. The result of b)) is shown. From this result, it can be seen that even with the same component content, the luminance differs for each measurement, and it is effective to estimate the component amount by creating a calibration curve for each measurement.

In the substance component estimation method and substance component estimation apparatus according to the embodiment of the present invention, the imaging unit 20 includes the scanner 23. However, the present invention is not limited to this, and a digital image is acquired. Possible devices such as a digital still camera and a film scanner may be used.
Further, the substance is not limited to the above-mentioned plants and soils, and it is needless to say that any substance may be targeted.
For example, any kind of plant, animal, organic matter, inorganic matter, etc. may be used. For example, food (including beverages), water (river water, tap water) and the like are assumed as substances.

It is a figure which shows the component estimation apparatus of the substance which concerns on embodiment of this invention. It is a figure which shows the extraction example of the container and image of the component estimation apparatus of the substance which concerns on embodiment of this invention. It is a figure which shows the tray of the component estimation apparatus of the substance which concerns on embodiment of this invention with the dimension example. It is the table | surface figure and graph which show the measured value and the estimated value 1 which concern on Example 1 of this invention. It is a table | surface figure and graph which show the measured value and the estimated value 2 which concern on Example 1 of this invention and made the measurement date different. It is a table | surface figure which shows the estimated value 3 estimated using the result of the previous measurement data as a calibration curve according to Example 1 of the present invention. It is a graph which shows the correlation of the data which concern on FIG. It is a graph which shows the correlation of the data which concern on FIG. It is a graph which shows the correlation of the data which concern on FIG. It is a table | surface which shows the measured value of a reference | standard sample in connection with Example 2 of this invention. It is a graph which shows the correlation of the data which concern on Example 2 of this invention. It is a table | surface figure which compares and shows the measured value in the different measurement time of a reference | standard sample concerning the experiment example of this invention. It is a table | surface figure which compares and shows the measured value in the different measurement day of a reference | standard sample concerning the experiment example of this invention. It is a figure which shows an example of the conventional component estimation apparatus of a substance.

Explanation of symbols

K sample C reference sample 1 sample container 2 reference container 3 bottom wall 10 tray 11 hole 20 imaging unit 21 base 23 scanner 30 calculation unit 31 reference image extraction unit 32 sample image extraction unit 33 reference luminance detection unit 33a pixel luminance detection unit 33b Average luminance calculating unit 34 Correlation calculating unit 35 Sample luminance detecting unit 35a Pixel luminance detecting unit 35b Average luminance calculating unit 36 Component amount calculating unit

Claims (13)

A specimen having a predetermined amount of substance or an extract of a predetermined amount of substance is accommodated in a specimen container, the surface of the specimen contained in the specimen container is imaged by an imaging unit, and an image captured by the imaging unit is obtained. In the component estimation method for a substance that detects the luminance of the surface of the specimen based on the detected luminance and calculates the component amount of a specific component among the contained components in the substance based on the detected luminance,
A plurality of reference specimens, each of which is a reference specimen in a form equivalent to the above-described specimen whose amount of a specific component is known in advance, each having a different amount of the known component, are housed in a plurality of reference containers. The imaging unit simultaneously images the surface of the sample accommodated in the sample container and the surface of the reference sample accommodated in each reference container, and each reference reference is based on the image captured by the imaging unit. Detecting the luminance of the surface of the reference sample contained in the container, calculating a correlation between the luminance of the surface of the reference sample and the amount of a specific component related to the reference sample based on the detected luminance, and performing the imaging The luminance of the surface of the sample contained in the sample container is detected based on the image captured by the unit, and the component amount of the specific component related to the sample is determined from the detected luminance of the surface of the sample and the correlation. Of the substance characterized by calculating Minute estimation method.   The sample is configured to have a predetermined amount of substance, and the reference sample is configured to have the same kind of substance as the predetermined amount of the sample whose amount of a specific component is known in advance. The method for estimating a component of a substance according to claim 1.   The specimen is composed of an extract of a predetermined amount of a substance, and the reference specimen is composed of a reagent in which a specific component is mixed and a component amount of the specific component is known in advance. The component estimation method of the substance of Claim 1.   4. The method for estimating a component of a substance according to claim 1, wherein the substance is a plant.   The method for estimating a component of a substance according to any one of claims 1 to 3, wherein the substance is soil. A specimen container that contains a specimen having a predetermined amount of substance or a predetermined amount of substance extract, an imaging section that images the surface of the specimen contained in the specimen container, and an image captured by the imaging section In the substance component estimation apparatus comprising: a detection unit that detects the luminance of the surface of the specimen based on the calculation unit that calculates a component amount of a specific component among the contained components in the substance based on the detected luminance;
A plurality of reference containers each containing a plurality of reference samples, each of which is a reference sample in a form equivalent to the above-described sample whose component amount of a specific component is known, and each of which has a different known component amount The imaging unit is configured to simultaneously image the surface of the specimen housed in the specimen container and the surface of the reference specimen contained in each of the reference containers;
Based on the luminance detected by the reference luminance detecting means, the reference luminance detecting means for detecting the luminance of the surface of the reference specimen contained in each reference container based on the image picked up by the imaging section, Correlation calculating means for calculating the correlation between the luminance of the surface of the reference sample and the component amount of the specific component related to the reference sample, and the sample container accommodated in the sample container based on the image captured by the imaging unit A sample luminance detecting unit for detecting the luminance of the surface of the sample, a luminance of the surface of the sample detected by the sample luminance detecting unit, and a correlation calculated by the correlation calculating unit; An apparatus for estimating a component of a substance, comprising: a component amount calculating means for calculating a component amount.   7. The substance component estimation apparatus according to claim 6, wherein the imaging unit is configured to image a plurality of the specimen containers. Using the sample container and the reference container formed of a transparent material, the imaging unit includes a base on which the sample container and the reference container are placed, the sample container on the base, and A scanner that scans the reference container and acquires an image through the bottom wall of each container;
The calculation unit includes reference image extraction means for extracting an image of a predetermined range of an image of only the surface of the reference specimen of the reference container from the image output from the scanner, and the image output from the scanner A sample image extracting means for extracting an image of a predetermined range that is an image of only the surface of the sample in the sample container,
The reference luminance detecting means of the calculating unit includes a pixel luminance detecting means for detecting the luminance of each pixel in the image extracted from the reference image extracting means, and an average from the luminance of each pixel detected by the pixel luminance detecting means. An average luminance calculating means for calculating luminance,
The sample luminance detection unit of the calculation unit includes a pixel luminance detection unit that detects the luminance of each pixel in the image extracted from the sample image extraction unit, and an average from the luminance of each pixel detected by the pixel luminance detection unit. An average luminance calculating means for calculating luminance,
The correlation calculation means is configured to calculate the correlation based on the average brightness for the reference specimen of each reference container calculated by the average brightness calculation means of the reference brightness detection means,
7. The component amount calculating unit is configured to calculate a component amount based on the average luminance of the sample in the sample container calculated by the average luminance calculating unit of the sample luminance detecting unit. 7. A component estimation apparatus for a substance according to 7.   A plurality of containers each having a bottom wall exposed and having a tray that can be placed on the base of the imaging unit; and the scanner is configured to scan the bottom walls of the plurality of containers provided on the tray. 9. The component estimation apparatus for a substance according to claim 8,   The sample is configured to have a predetermined amount of substance, and the reference sample is configured to have the same kind of substance as the predetermined amount of the sample whose amount of a specific component is known in advance. The substance component estimation apparatus according to any one of claims 6 to 9,   The specimen is composed of an extract of a predetermined amount of a substance, and the reference specimen is composed of a reagent in which a specific component is mixed and a component amount of the specific component is known in advance. The component estimation apparatus of the substance in any one of Claims 6 thru | or 9.   12. The substance component estimation apparatus according to claim 6, wherein the substance is a plant.   The apparatus for estimating a component of a substance according to claim 6, wherein the substance is soil.