WO2019102660A1

WO2019102660A1 - Odor detection apparatus and program

Info

Publication number: WO2019102660A1
Application number: PCT/JP2018/029726
Authority: WO
Inventors: 秋山　博
Original assignee: コニカミノルタ株式会社
Priority date: 2017-11-27
Filing date: 2018-08-08
Publication date: 2019-05-31
Also published as: JPWO2019102660A1; JP7070586B2

Abstract

The present invention distinguishes odor types. This odor detection apparatus identifies an odor component (nonenal, diacetyl, isovaleric acid, etc.) included in air being measured and the concentration of said odor component on the basis of output values from a plurality of odor sensors having mutually different characteristics for reacting to odors (step S7), distinguishes the odor type (elderly body odor, odor of middle-aged overweight person, sweat, etc.) on the basis of the identified odor component and concentration of said odor component (step S8), and outputs the distinguished odor type (step S11). Specifically, a value corresponding to an odor component identified in advance and the concentration of said odor component is subtracted from the output values of the plurality of odor sensors (step S10), and an odor component included in the air being measured and the concentration of said odor component are identified on the basis of the subtracted results, these steps being repeated to identify, in order, a plurality of odor components included in the air being measured and the concentrations thereof.

Description

Odor detection device and program

The present invention relates to an odor detector and program.

Generally, odor is sensory, and it is difficult to objectively recognize the impression or degree that human beings receive from odor. Many people are concerned about odors that are generally considered to be odors, such as body odor and bad breath, but it is difficult to check their own odor with human olfactory sense.

In order to measure halitosis quantitatively, for example, an exhalation component measuring device is used which measures the concentration of gas contained in human exhalation by a gas sensor whose output value changes according to the concentration of gas to be detected. (See Patent Document 1). This device is equipped with one gas sensor inside.

JP, 2011-232058, A

However, since the gas sensor does not react only to a specific detection target gas, and only the reaction characteristics differ depending on the gas, including the reaction intensity, the measurement using one sensor as in the above-mentioned prior art In the device, a situation occurs in which only the intensity of the odor can be detected because it responds to both a good odor and a bad odor. For example, there is a problem that the type of odor can not be determined because, even if it is trying to measure halitosis, it responds similarly to the odor of dumplings or the odor of toothpaste.

The present invention has been made in view of the above-mentioned problems in the prior art, and it is an object of the present invention to determine the type of odor.

In order to solve the above problems, the invention according to claim 1 is an odorant included in a gas to be measured based on output values of a plurality of odor sensors different from one another in characteristics that respond to odor and the plurality of odor sensors. The apparatus comprises: identification means for identifying the component and the concentration thereof; determination means for determining the type of the odor based on the identified odorant component and the concentration thereof; and output means for outputting the determined type of the odor. The odor detection device, wherein the identification means subtracts the odor component and the value corresponding to the odor component that have already been identified from the output values of the plurality of odor sensors, and based on the result of the subtraction, the measurement target A plurality of odorant components contained in the gas to be measured and their concentrations are prioritized by repeating specifying the odor components contained in the gas and their concentrations. To.

The second aspect of the present invention is the odor detection device according to the first aspect, wherein the output of the plurality of odor sensors when the odor component of the concentration is targeted for each concentration of each of the plurality of odor components A storage unit for storing a classifier that is a machine learning result obtained by machine learning in advance using the combination of values as an input and the odor component and the concentration thereof as an output; and the identification unit uses the identifier to execute the identification The odor component contained in the gas to be measured and its concentration are specified.

The invention according to claim 3 is the odor detection device according to claim 2, wherein a combination of output values of the plurality of odor sensors when a predetermined odor is targeted is input to the discriminator. When the predetermined means includes the machine learning result which is machine-learned in advance using the predetermined odor as the output, and the identification means identifies that the predetermined odor is included in the gas to be measured using the discriminator, A value corresponding to the predetermined odor is subtracted from the output values of the plurality of odor sensors, and the result of the subtraction is input to the discriminator to specify the odorant component contained in the gas to be measured and the concentration thereof.

The invention as set forth in claim 4 is the odor detection device according to any one of claims 1 to 3, wherein the discrimination means is based on the specified odorant component and the concentration thereof. The intensity is determined, and the output means outputs an intensity of the type of the determined odor.

The invention according to claim 5 is the odor detection device according to claim 4, wherein the second storage means stores a table in which the type of odor and the intensity thereof are associated for each concentration of each of a plurality of odor components. The discrimination means discriminates the type and intensity of the odor corresponding to the identified odorant component and its concentration using the table.

According to the sixth aspect of the present invention, in the odor detection device according to the fourth or fifth aspect, the output means causes the display means to display the intensity of the type of the determined odor.

According to the seventh aspect of the invention, in the odor detection device according to any one of the first to sixth aspects, the output means causes the display means to display the type of the determined odor.

According to an eighth aspect of the present invention, in the odor detection device according to the seventh aspect, the output means is configured to determine the odor component corresponding to the type of the determined odor and the order in which the concentration thereof is specified. The type of the determined odor is displayed so that the order of the type of the odor can be known.

The invention described in claim 9 is the odor detection device according to any one of claims 6 to 8, wherein the display means is a display means provided in a smartphone.

According to the invention as set forth in claim 10, in the odor detecting device according to any one of claims 1 to 9, the discriminating means is one of an age odor, a middle fat odor and a sweat odor as a type of the odor. Determine at least one.

The invention according to claim 11 is a computer according to an acquisition means for acquiring output values from a plurality of odor sensors having different characteristics responsive to odor, and a measurement target based on the acquired output values of the plurality of odor sensors An identifying unit for identifying an odorant component contained in a gas and its concentration, a discriminating unit for discriminating the type of odor based on the identified odorant component and its concentration, an output unit for outputting the discriminated odor type; The identification unit subtracts the odor component already identified from the output values of the plurality of odor sensors and a value corresponding to the concentration thereof, and based on the result of the subtraction, By repeatedly specifying the odorant components contained in the gas to be measured and the concentrations thereof, a plurality of nios contained in the gas to be measured can be obtained. The components and their concentrations, to identify with a rank.

According to the present invention, the type of odor can be determined.

It is a block diagram of an odor detection system. It is an example of the odor type discrimination table. It is a figure showing an outline of smell detection processing. It is an example of the time change of the output value of an odor sensor. It is an example of the time change of the output value of the odor sensor corresponded to the identified odor component and its density | concentration. It is the result of subtracting the value shown in FIG. 4B from the value shown in FIG. 4A. It is a flowchart which shows the odor detection process performed in an odor detection apparatus. It is an example of the measurement result display screen displayed on a smart phone.

Hereinafter, with reference to the drawings, an embodiment of an odor detector according to the present invention will be described. The present invention is not limited to the illustrated example.

The configuration of the odor detection system 100 is shown in FIG. The odor detection system 100 is configured by the odor detection device 10 and the smartphone 20. The odor detection device 10 and the smartphone 20 have a near field wireless communication function, and are capable of mutual communication by Bluetooth (registered trademark).

The odor detection device 10 includes a control unit 11, four

odor sensors

12A, 12B, 12C, 12D, an ADC (Analog to Digital Converter) 13, a storage unit 14, an operation switch 15, a battery 16, and a communication unit. It has 17 mag.

The control unit 11 is configured by a CPU (Central Processing Unit) or the like, and centrally controls the processing operation of each unit of the odor detection device 10. Specifically, the control unit 11 reads various processing programs stored in the storage unit 14 and performs various processes in cooperation with the programs.

The

odor sensors

12A, 12B, 12C, and 12D are semiconductor gas sensors having different odor-responsive properties, such as what odor is strongly responsive. The

odor sensors

12A, 12B, 12C, 12D convert the concentration of the gas to be detected (the gas to be mainly reacted) into an electrical quantity, and output an electric signal corresponding to the gas concentration. As the

odor sensors

12A, 12B, 12C, 12D, for example, gas sensors for detecting VOC (volatile organic compounds), gas sensors for detecting CO, gas sensors for detecting hydrogen, gas sensors for detecting hydrocarbons, alcohol detection Gas sensors, cigarette detection gas sensors, etc. are used. Each

odor sensor

12A, 12B, 12C, 12D reacts to a plurality of odorant components, and does not react to only one odorant component. The odor component is a chemical substance that constitutes odor. Further, the

odor sensors

12A, 12B, 12C, 12D may be MEMS (Micro Electro Mechanical Systems) type sensors.
The ADC 13 converts analog signals output from the

odor sensors

12A, 12B, 12C, and 12D into digital signals, and outputs the digital signals to the control unit 11.

The storage unit 14 is configured by a non-volatile semiconductor memory or the like, and stores various processing programs, parameters, files and the like necessary for executing the programs. A discriminator 141 and an odor type discrimination table 142 are stored in the storage unit 14 (storage means, second storage means).

The discriminator 141 is a machine learning result generated in advance by an external device at the development stage of the odor detection device 10. At the time of machine learning, each of a plurality of odorant components (chemical substances) is prepared for each concentration. In the external device, the output value (waveform) of the

odor sensor

12A, 12B, 12C, 12D and the same odor sensor (hereinafter referred to as odor sensor for learning) are acquired for each concentration of each of a plurality of odor components, A classifier 141 is generated by performing machine learning with the combination of the output value of each odor sensor for learning when the odor component of the concentration is a target, and the odor component and the concentration thereof as an output. That is, a combination of output values of a plurality of learning odor sensors is input, and a set of data having an odor component and its density as an output is machine learning as teacher data.

The “same kind of odor sensor” is a sensor having the same characteristics as the

target odor sensors

12A, 12B, 12C, 12D (that can obtain the same output), for example, sensors of the same model number.
As machine learning, for example, a neural network, in particular, a learning vector quantization (LVQ) neural network is used. For example, the discriminator 141 is constructed using, as a feature value, the rise of a waveform obtained by plotting the gas concentration which is the output value of the learning odor sensor along the elapsed time from the start of measurement, the peak value, etc. ing.

Examples of the odorant component include components that cause malodor such as nonenal, diacetyl, isovaleric acid, ammonia and the like. Nonenal is a component that causes aging odor (body odor generated with aging). Diacetyl is a component that causes middle fat odor (a greasy body odor often found in middle-aged men). Isovaleric acid is a component that causes sweat odor (body odor due to sweat).

Also, the classifier 141 has a combination of output values of a plurality of learning odor sensors for a predetermined odor as an input, and includes the machine learning result in which the predetermined odor is output as machine learning. Be Examples of the predetermined odor include perfumes, shampoos, scents of softeners, etc., which are generally felt as good odors.

Here, when the classifier 141 is generated in advance, the odor sensor for learning is used, but the machine using the

odor sensors

12A, 12B, 12C, 12D itself actually mounted on the odor detection device 10 It may be made to learn.

The odor type discrimination table 142 is a table in which the types of odor and their intensities (levels) are associated with each other for each of a plurality of odor components. The odor type discrimination table 142 is information serving as a determination standard for converting the concentration of each odorant component into the type (and the intensity thereof) of odor.
The kind of odor is what the human being classified and named to the odor having some characteristics. Types of odor include age-related odor, middle oily odor and sweat odor.

For example, the odor type discrimination table 142 is prepared for each odor component.
FIG. 2 shows an example of the odor type discrimination table 142 of diacetyl. In FIG. 2, the level (1 to 10) of "middle fat odor" is associated with each concentration of the odorant component "diacetyl". In addition, the definition method of the level of "middle greasy smell" is described in FIG. 2 as reference. Each level is determined by sensory evaluation based on human sense of smell, and is similar to human feeling. For example, the level “1” of “middle fat odor” is an intensity that causes odor when smelled at a distance of 1 cm.
In the odor type discrimination table 142 of nonenal, the level of “aging odor” is associated with each concentration of the odor component “nonenal”.
In the odor type discrimination table 142 of isovaleric acid, the level of “sweat odor” is associated with each concentration of the odorant component “isovalerate”.

Further, the storage unit 14 stores in advance the output values (waveforms) of the plurality of

odor sensors

12A, 12B, 12C, 12D when targeting the odor component of the concentration for each concentration of each odor component. There is.
Further, in the storage unit 14, output values (waveforms) of a plurality of

odor sensors

12A, 12B, 12C, 12D when predetermined odors are targeted are stored in advance.
The data of the output value corresponding to each concentration of each odorant and the data of the output value corresponding to the predetermined odor stored in the storage unit 14 are obtained in advance using a learning odor sensor. In addition, it may be obtained in advance using the

odor sensors

12A, 12B, 12C, 12D themselves actually mounted on the odor detection device 10.

The operation switch 15 is configured of a power switch for turning on / off the power, a measurement switch for instructing measurement start, and the like, and outputs an operation signal according to pressing of the operation switch 15 by the user to the control unit 11.
The battery 16 supplies power to each part of the odor detection device 10. As the battery 16, a removable dry battery, a rechargeable battery or the like is used.

The communication unit 17 has an interface for performing data communication with the smartphone 20 by Bluetooth wireless communication, and transmits / receives data to / from the smartphone 20 according to the communication standard of BLE (Bluetooth Low Energy).

FIG. 3 shows an outline of the odor detection process. In the present invention, the first step of identifying the odorant component and its concentration from the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D, and the second step of discriminating the kind of odor from the odorant component and its concentration Process separately.

<First stage>
The control unit 11 (specifying means) determines odor components (nonenal, diacetyl, isovaleric acid, etc.) contained in the gas to be measured based on the output values (waveforms) of the plurality of

odor sensors

12A, 12B, 12C, 12D Identify the concentration. Specifically, the control unit 11 inputs the output values acquired from the plurality of

odor sensors

12A, 12B, 12C, 12D to the gas to be measured into the discriminator 141, and the odorant component output from the discriminator 141 And the concentration thereof is specified as an odorant component contained in the gas to be measured and its concentration.

<Step 1 Exception>
The control unit 11 specifies that the predetermined odor is included in the gas to be measured, when a predetermined odor (a good odor such as a perfume, a shampoo, or a softener) is obtained as an output of the discriminator 141.

<Repeat of the first stage>
The identification of the odor component and the concentration thereof by the discriminator 141 and the identification of the predetermined odor are to find the strongest odor component or the predetermined odor (1st place) in the gas to be measured. Therefore, it is necessary to specify an odor component at position 2 or higher or a predetermined odor without the influence of the odor component identified earlier or the predetermined odor.

The control unit 11 subtracts the odor component already specified and the value corresponding to the concentration from the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D, and based on the result of the subtraction, A plurality of odorant components contained in the gas to be measured and their concentrations are prioritized and identified by repeating the identification of the contained odorous components and the concentration thereof.

First, the control unit 11 subtracts the value corresponding to the odor component identified first and its density from the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D. Specifically, when the control unit 11 targets only the specified odorous component of the concentration specified first, the output values to be output from the plurality of

odor sensors

12A, 12B, 12C, 12D (each The temporal change of the odor sensor is read out from the storage unit 14, and the read out values are subtracted from the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D.

An example of the time change of the output value of the odor sensor 12A when the odor component and its concentration are identified from the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D is shown in FIG. 4A and identified An example of the time change of the output value of the odor sensor 12A corresponding to the odor component and the concentration thereof is shown in FIG. 4B. The result of subtracting the value shown in FIG. 4B from the value shown in FIG. 4A is shown in FIG. 4C at each time shown in the horizontal axis.
The output values of the

odor sensors

12B, 12C and 12D are similarly processed.

The control unit 11 is included in the gas to be measured based on the result of subtraction (the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D excluding the first identified odor component and the influence of the concentration thereof). Identify the second odorant and its concentration. Specifically, the control unit 11 inputs the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D excluding the influence of the first identified odor component and its concentration to the discriminator 141, and the discriminator The odorant component output from 141 and its concentration are specified as the second odorant component and its concentration. Then, from the output values of the plurality of

odor sensors

12A, 12B, 12C, and 12D used when specifying the second odor component, the control unit 11 corresponds to the second identified odor component and the concentration thereof. Subtract.

The control unit 11 determines a gas to be measured based on the result of subtraction (the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D excluding the effects of the first and second identified odor components and their concentrations). Identify the third odorant contained in and its concentration. Specifically, the control unit 11 inputs to the discriminator 141 the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D excluding the effects of the first and second identified odor components and their concentrations. The odor component output from the discriminator 141 and its concentration are specified as the third odor component and its concentration.

As described above, the control unit 11 repeats the specification of the odorant component and the concentration while subtracting the odorant component specified from the output values of the plurality of

odor sensors

12A, 12B, 12C, and 12D and the value corresponding to the concentration. Thus, a plurality of odorant components contained in the gas to be measured and their concentrations are ranked and specified.

When the control unit 11 uses the discriminator 141 to specify that the gas to be measured contains a predetermined odor, each of the plurality of

odor sensors

12A, 12B, 12C, and 12D when the predetermined odor is targeted. The output value to be output is read from the storage unit 14, and the read values are subtracted from the output values of the plurality of

odor sensors

12A, 12B, 12C, and 12D. Then, the control unit 11 inputs the result of the subtraction to the discriminator 141, and specifies the odorant component contained in the gas to be measured and the concentration thereof.

Second stage
The control unit 11 (determination means) determines the type of odor based on the identified odorant component and the concentration thereof. The control unit 11 determines at least one of an aging odor, a middle fat odor, and a sweat odor as the type of odor. Furthermore, the control unit 11 determines the intensity (level) of the type of odor based on the identified odorant component and its concentration. Specifically, the control unit 11 determines, for each of the plurality of specified odor components, the type of odor component corresponding to the odor component and the concentration thereof, and the intensity thereof, using the odor type discrimination table 142.

The control unit 11 (output means) outputs the determined odor type. Furthermore, the control unit 11 outputs the intensity of the type of the determined odor. Specifically, the control unit 11 transmits to the smartphone 20 via the communication unit 17 in order to cause the display unit 22 of the smartphone 20 to display the type and the intensity of the determined odor. The control unit 11 displays the type of the determined odor based on the order in which the determined type of the odor is identified based on the odor component corresponding to the determined type of odor and the concentration thereof, and the display unit of the smartphone 20 Display on 22 The control unit 11 may generate screen display data in which the strength and the order of the determined odor types are known, and may transmit the data to the smartphone 20. Alternatively, on the smartphone 20 side, based on the type of odor transmitted from the odor detection device 10 and the strength or the rank thereof, the rank of the type of odor may be displayed.

As shown in FIG. 1, the smartphone 20 includes a control unit 21, a display unit 22, an operation unit 23, a first communication unit 24, a second communication unit 25, a storage unit 26, a speaker 27, a microphone 28, and the like.

The control unit 21 is configured by a CPU or the like, and centrally controls the processing operation of each unit of the smartphone 20. Specifically, the control unit 21 reads various processing programs stored in the storage unit 26, and performs various processing in cooperation with the programs.

The display unit 22 is configured by an LCD (Liquid Crystal Display) or the like, and displays various screens in accordance with an instruction of a display signal input from the control unit 21.
The operation unit 23 includes an operation key and a touch panel stacked on the display unit 22. The operation unit 23 outputs, to the control unit 21, an operation signal corresponding to the operation key and an operation signal according to the position of a touch operation by a finger of the user.

The first communication unit 24 wirelessly connects to a communication network including a mobile communication network via a base station or an access point, and communicates with an external device connected to the communication network.
The second communication unit 25 has an interface for performing data communication with the external device such as the odor detection device 10 by Bluetooth wireless communication. The second communication unit 25 transmits and receives data to and from an external device in accordance with the BLE communication standard.

The storage unit 26 is configured of a non-volatile semiconductor memory or the like, and stores various processing programs, parameters, files, and the like necessary for executing the programs. In the storage unit 26, an odor detection application program (hereinafter referred to as an odor detection application) for performing odor detection using the odor detection device 10 is installed.

The speaker 27 converts an electrical signal received from an external device through the first communication unit 24 into an audio signal, and outputs audio.
The microphone 28 detects a sound wave, converts it into an electric signal, and outputs the electric signal to the control unit 21 or the first communication unit 24.

The control unit 21 causes the display unit 22 to display the type and the intensity of the odor received from the odor detection device 10 via the second communication unit 25 in cooperation with the odor detection application.

Next, the operation of the odor detection system 100 will be described.
FIG. 5 is a flowchart showing the odor detection process performed in the odor detection device 10. This processing is realized by software processing by cooperation of the control unit 11 and the program stored in the storage unit 14. In addition, as a premise of the process, the user activates the odor detection application in the smartphone 20 by the operation from the operation unit 23, and turns on the odor detection device 10.

First, when the user presses the operation switch 15 (specifically, the measurement switch) (step S1), the control unit 11 detects that the operation switch 15 is pressed. The user holds the odor detection device 10 at a site (head, ears, aside, feet, etc.) where the body odor of the user's own body is to be measured, and fixes the position of the odor detection device 10 for a fixed time.

The control unit 11 (acquisition unit) is configured to control the

gas sensor

12A, 12B, 12C, 12D with respect to the gas to be measured, which exists near the position where the user holds the odor detection device 10 for a fixed time after the operation switch 15 is pressed. An output value is acquired (step S2).

Next, the control unit 11 determines whether the gas to be measured is odorless based on the output values obtained from the plurality of

odor sensors

12A, 12B, 12C, 12D (step S3). For example, when the output values of all the

odor sensors

12A, 12B, 12C, and 12D are less than a predetermined threshold value, the control unit 11 determines that there is no significant odor input, that is, odorless.

When it is determined that the gas to be measured is not odorless (step S3; NO), the control unit 11 stores the combination of the output values obtained from the plurality of

odor sensors

12A, 12B, 12C, and 12D in the storage unit 14. Input to the stored discriminator 141 (step S4).

Here, the control unit 11 determines whether the result of the predetermined odor (perfume, shampoo, softener, etc.) is obtained as the output from the discriminator 141 or the odor component and the concentration thereof are obtained ( Step S5).

When the result of the predetermined odor is obtained as the output from the discriminator 141 (step S5; predetermined odor), the control unit 11 outputs the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D (immediately before In step S4, the value corresponding to the predetermined odor is subtracted from the combination of the values input to the discriminator 141 (step S6). Then, the process returns to step S4, the control unit 11 inputs the result of the subtraction to the discriminator 141, and repeats the process.

In step S5, when an odorant component and its concentration are obtained as an output from the discriminator 141 (step S5; odorant component and concentration), the control unit 11 determines the odorant component and its concentration as a gas to be measured. Are specified as odorous components and their concentrations, and the specified odorous components and their concentrations are ranked and stored in the storage unit 14 (step S7).

Next, the control unit 11 refers to the odor type discrimination table 142 stored in the storage unit 14 to discriminate the specified odor component and the type and intensity of odor corresponding to the concentration thereof (step S8). .

Here, the control unit 11 determines whether or not the odor component that is the cause of the offensive odor has been identified up to the third from the gas to be measured (step S9).

When the odor component has not been identified up to the third (step S9; NO), the control unit 11 outputs the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D (in the previous step S4, the identifier 141 A value corresponding to the specified odorant component and its concentration is subtracted from the combination of input values) (step S10). Then, the process returns to step S4, the control unit 11 inputs the result of the subtraction to the discriminator 141, and repeats the process.

When the odor component is specified up to the third in step S9 (step S9; YES), or when it is determined in step S3 that the gas to be measured is odorless (step S3; YES), the control unit 11 Transmits the odor detection result to the smartphone 20 via the communication unit 17 (step S11). Specifically, when the type of odor and the intensity thereof are determined, the control unit 11 transmits the type of odor and the intensity thereof to the smartphone 20 in order to cause the display unit 22 of the smartphone 20 to display. More specifically, the control unit 11 determines the odor component corresponding to the determined odor type and the order in which the concentration is identified as the order of the type of the odor, and the intensity and the order of the determined odor type are Data for screen display that can be understood may be generated and transmitted to the smartphone 20. Alternatively, the control unit 11 may transmit the order of the determined odor type to the smartphone 20 in association with the odor type. If it is determined that the odor is odorless, the control unit 11 transmits, to the smartphone 20, the result that the body odor is not detected (the level "0" for each type of odor).
Thus, the odor detection process is completed.

In the smartphone 20, the odor detection result transmitted from the odor detection device 10 is displayed on the display unit 22. Specifically, the type and the intensity of the detected odor are displayed on the display unit 22 so that the order of the types of the odor can be known.

FIG. 6 shows an example of the measurement result display screen 30 displayed on the display unit 22 of the smartphone 20. In the measurement result display screen 30, there are a first odor type display area 31, a second odor type display area 32, a third odor type display area 33, a message display area 34, a "rescale" button 35, and a "completion" button 36. included.

The first odor type display area 31 includes an odor type display area 31A and odor

level display areas

31B and 31C. In the odor type display area 31A, the type of odor (the type of odor at the first place) corresponding to the first identified odor component is displayed. In the odor

level display areas

31B and 31C, the intensity of the type of odor displayed in the odor type display area 31A is displayed in 10 levels. In the odor level display area 31B, the intensity is graphed and displayed, and in the odor level display area 31C, the intensity is displayed as a numerical value. In the example of FIG. 6, "Sweat odor" is displayed as the type of odor in the odor type display area 31A, and it is displayed in the odor

level display areas

31B and 31C that the level of the sweat odor is "5" in 10 steps. It is done.

In the second odor type display area 32, the type of odor (type of odor at the second place) and the intensity corresponding to the second identified odor component are displayed.
In the third odor type display area 33, the type of odor (type of odor at the third place) and the intensity corresponding to the third identified odor component are displayed.

In the message display area 34, explanations, advice, etc. for the measurement result are displayed.
The “re-measure” button 35 is a button for instructing re-measure.
The “completion” button 36 is a button for instructing measurement completion.

As described above, according to the present embodiment, the gas to be measured is included in the gas to be measured while removing the influence of the odorant component and its concentration which have already been specified from the output values of the plurality of

odor sensors

12A, 12B, 12C, 12D. A plurality of odorant components and their concentrations can be prioritized and identified by repeating identification of the odorant components and their concentrations. In this way, it is possible to specify not only the first strongest (more) odor component but also the second and third odor components. Therefore, the plurality of types of odor can be determined based on the plurality of identified odorant components and their concentrations. For example, as the type of odor, it is possible to determine an anxious body odor such as aging odor, middle fat odor, sweat odor and the like. The user can use the odor detection device 10 to objectively check his or her odor to prevent the surrounding people from feeling uncomfortable.

Specifically, an odor component and its concentration can be specified using a machine learning result (classifier 141) that has been machine-learned in advance.

Further, not only the type of odor but also the intensity of the type of odor is determined based on the odorant component contained in the gas to be measured and the concentration thereof, the user can recognize the degree of the quantified odor. .
Specifically, by using the odor type discrimination table 142, it is possible to easily discriminate between the specified odor component and the type and intensity of odor corresponding to the concentration.

In summary, in the second step, each odorant component is specified by specifying how much (concentration) each odorant component (nonenal, diacetyl, isovaleric acid, etc.) is contained in the gas to be measured in the first step. The “kind of odor” that is the main cause can be determined, and the concentration of each odorant component is an index for determining the strength of the “type of odor”. That is, by specifying the concentration of each odorant component, it becomes possible to quantify the "type of odor".
At the first stage, odorant components that are directly linked to the “type of odor” such as aging odor, middle fat odor, sweat odor, etc. to be finally determined, ie, the odorant component that is the main cause of each “type of odor” It is important to identify the concentration.

Also, predetermined odor that is generally regarded as good odor can be excluded from the detection of malodor such as body odor by performing machine learning so as to specify that the gas to be measured includes the predetermined odor. .

Further, since the type and the intensity of the determined odor are displayed on the display unit 22 of the smartphone 20, the user can be notified of the odor detection result. In particular, the user can be notified of an odor to be noted by displaying the plurality of odor types so as to indicate the order.

The description in the above embodiment is an example of the odor detector according to the present invention, and the present invention is not limited to this. The detailed configuration and the detailed operation of each part constituting the apparatus can be appropriately modified without departing from the scope of the present invention.

For example, in the above embodiment, the type and intensity of the odor determined by the odor detection device 10 are displayed on the display unit 22 of the smartphone 20. However, the odor detection device 10 itself includes the display portion, and the odor is The type and the intensity of the odor may be displayed on the display unit of the detection device 10 (output means). At this time, the type of the determined odor may be displayed based on the order in which the type of the odor is identified based on the order in which the odor component corresponding to the determined type of odor and the concentration thereof are specified. Alternatively, the type and the intensity of the determined odor may be indicated by lighting of an LED (Light Emitting Diode) or the like included in the odor detection device 10.

In the above embodiment, the odor detector 10 determines not only the type of odor but also the intensity of the type of odor based on the specified odor component and its concentration. If the concentration of the component is equal to or higher than a predetermined threshold value, it may be determined that the “type of odor” corresponding to the odor component is detected.

Alternatively, the discriminator 141 and the odor type discrimination table 142 may be prepared in advance for each measurement site such as the head, ears, aside, and feet. In this case, when the user measures the body odor, the measurement site may be designated from the odor detection device 10 or the smartphone 20 to switch to the determination reference (classifier 141 and the odor type discrimination table 142) corresponding to the measurement site. .

Further, in the above embodiment, when the third odorant component is identified, the identification of the odorant component by the discriminator 141 is ended, but the number of the identified odorant components is not limited to this. Also, the output value of the plurality of

odor sensors

12A, 12B, 12C, 12D is subtracted the value corresponding to the specified odor component and its concentration, or the value corresponding to the specified predetermined odor, and the result is subtracted. It is possible to terminate the specification of the odorant component when the value of T falls below a predetermined threshold value.

The odor detector and program according to the present invention may be used in the technical field of measuring odors such as human body odor.

DESCRIPTION OF SYMBOLS 10 odor detection apparatus 11

control part

12A, 12B, 12C, 12D odor sensor 14 memory | storage part 15 operation switch 17 communication part 20 smart phone 21 control part 22 display part 23 operation part 25 2nd communication part 26 memory | storage part 100 odor detection system 141 identification Device 142 odor type discrimination table

Claims

Multiple odor sensors with different odor-responsive characteristics,
Specifying means for specifying an odorant component contained in the gas to be measured and its concentration based on output values of the plurality of odor sensors;
Discrimination means for discriminating the type of odor based on the specified odor component and its concentration;
Output means for outputting the type of the determined odor;
An odor detection device comprising
The specifying means subtracts the odor component and the value corresponding to the concentration from the output values of the plurality of odor sensors, and based on the result of subtraction, the odor component contained in the gas to be measured and The odor detection device which ranks and specifies a plurality of odor components contained in the gas to be measured and their concentrations by repeating specifying the concentration.
For each concentration of a plurality of odorant components, a combination of the output values of the plurality of odor sensors when the odorant component of the concentration is a target is input, and the odorant component and its concentration are output as machine learning Storage means for storing a discriminator that is a machine learning result obtained by
The odor detection device according to claim 1, wherein the identification unit identifies the odor component contained in the gas to be measured and the concentration thereof using the discriminator.
The discriminator includes machine learning results in which machine learning is performed in advance by using a combination of output values of the plurality of odor sensors for a predetermined odor as an output and the predetermined odor as an output.
The specifying means subtracts the value corresponding to the predetermined odor from the output value of the plurality of odor sensors when specifying that the predetermined gas is included in the gas to be measured using the discriminator. The odor detection device according to claim 2, wherein the subtraction result is input to the discriminator to specify the odorant component contained in the gas to be measured and the concentration thereof.
The determination means determines the intensity of the type of the odor based on the identified odorant component and the concentration thereof.
The odor detection device according to any one of claims 1 to 3, wherein the output means outputs an intensity of the type of the determined odor.
A second storage unit configured to store a table in which the type of odor and the intensity thereof are associated with each other for each of a plurality of odor components;
5. The odor detection device according to claim 4, wherein the discrimination means discriminates the type and intensity of the odor corresponding to the identified odor component and its concentration using the table.
The odor detection device according to claim 4 or 5, wherein the output means causes the display means to display the intensity of the type of the determined odor.
The odor detection device according to any one of claims 1 to 6, wherein the output unit causes the display unit to display the type of the determined odor.
The output means displays the type of the determined odor so that the order of the type of the odor can be known based on the order in which the odor component corresponding to the determined type of the odor and the concentration thereof are specified. The odor detection device according to Item 7.
The odor detection device according to any one of claims 6 to 8, wherein the display means is a display means provided in a smartphone.
The odor detector according to any one of claims 1 to 9, wherein the discrimination means discriminates at least one of an age odor, a middle oil odor and a sweat odor as a type of the odor.
Computer,
Acquisition means for acquiring output values from a plurality of odor sensors having different characteristics that respond to odor,
Specifying means for specifying an odorant component contained in the gas to be measured and its concentration based on the output values of the plurality of acquired odor sensors;
Discrimination means for discriminating the type of odor based on the identified odorant component and the concentration thereof
Output means for outputting the type of the determined odor;
A program to function as
The specifying means subtracts the odor component and the value corresponding to the concentration from the output values of the plurality of odor sensors, and based on the result of subtraction, the odor component contained in the gas to be measured and A program for specifying and specifying a plurality of odorant components contained in the gas to be measured and their concentrations by repeating specifying the concentration.