JP2018134051A - Information processing device, information processing method and information processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information processing device, an information processing method and an information processing program that enable an organism present in an environment to be can easily identified.SOLUTION: A learning unit performs machine learning on the basis of training data in which environmental information and biological information are caused to correspond to each other, the environmental information including at least a captured image obtained by imaging an environment where an organism is present, the biological information being the result of identifying the organism present in the environment on the basis of the base sequence. The image acquisition unit acquires the captured image obtained by imaging the environment. The organism identification unit identifies the organism corresponding to the captured image acquired by the acquisition unit on the basis of the result of machine learning performed by the learning unit.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an information processing apparatus, an information processing method, and an information processing program.

  Conventionally, it has been proposed to identify microorganisms from an image taken through a microscope using artificial intelligence or image recognition technology. For example, Patent Document 1 describes a biota diagnosis support system that takes in image data of treated water imaged by a microscope camera and identifies and counts microorganisms from the image data by image recognition processing using a model base and a matching method. Yes. Non-Patent Document 1 describes a method of classifying various fungi from a microscope image using an artificial neural network (ANN: Artificial Neural Network, hereinafter also simply referred to as a neural network or NN).

Japanese Patent Application Laid-Open No. 8-97084

Ayse Elif Ozturk, Sinan Alkan, and Celaleddin Ozturk, Classification of Fungus Spore Images Using Ridgelet Transform and ANN, Proceedings of the International Conference on Machine Vision and Machine Learning, Paper No. 130, August 14-15, 2014

  However, conventional inspection methods using images cannot be applied unless the morphological features of the organism to be identified are visible. In order to identify unrecognizable organisms using images, for example, an enlarged image is obtained using an optical microscope, or organisms are collected from the environment where these organisms exist and cultured. It takes a thing. Therefore, those organisms could not be easily identified.

  The present invention has been made in view of the above points, and an object of the present invention is to provide an information processing apparatus, an information processing method, and a program that can easily identify an organism present in the environment. .

  The present invention has been made to solve the above-described problems, and one aspect of the present invention provides environmental information including at least a captured image obtained by imaging an environment in which living organisms exist, and base sequences of organisms present in the environment. A learning unit that performs machine learning based on teacher data in association with biological information that is the result of identification based on the image, an image acquisition unit that acquires a captured image obtained by capturing an environment, and machine learning by the learning unit. And an organism identification unit that identifies an organism corresponding to the captured image acquired by the image acquisition unit based on the result.

  One embodiment of the present invention is an information processing method in an information processing apparatus, and includes environmental information including at least a captured image obtained by capturing an environment in which living organisms exist, and a result of identifying organisms existing in the environment based on a base sequence A learning step for performing machine learning based on teacher data in association with biological information, an image acquisition step for acquiring a captured image obtained by capturing an environment, and a result of machine learning by the learning step, A biological identification step of identifying a biological entity corresponding to the captured image acquired by the image acquisition step.

  According to one embodiment of the present invention, a computer corresponds to environmental information including at least a captured image obtained by imaging an environment in which an organism exists and biological information that is a result of identifying the organism present in the environment based on a base sequence. A learning step of performing machine learning based on the attached teacher data, an image acquisition step of acquiring a captured image of the environment, and an imaging acquired by the image acquisition step based on a result of machine learning by the learning step This is a program for executing an organism identification step for identifying an organism corresponding to an image.

  According to the present invention, organisms existing in the environment can be easily identified.

It is a conceptual diagram which shows the outline | summary of 1st Embodiment. 1 is a schematic block diagram illustrating a configuration of a computer system according to a first embodiment. It is a schematic block diagram which shows the information processing apparatus function structure which concerns on 1st Embodiment. It is a figure which shows the structural example of the neural network which concerns on 1st Embodiment. It is a figure which shows the example of the teacher data which concern on 1st Embodiment. It is a figure which shows an example of the initial integration which concerns on 1st Embodiment. It is a figure which shows an example of the latter period integration which concerns on 1st Embodiment. It is a figure which shows the other example of the late integration which concerns on 1st Embodiment. It is a schematic block diagram which shows the structure of the information processing system which concerns on 2nd Embodiment. It is a figure showing an example of medicine data concerning a 2nd embodiment. It is explanatory drawing which shows the transmission time of the biological information which concerns on 2nd Embodiment. It is a figure which shows an example of the control data which concerns on the modification of 2nd Embodiment. It is a schematic block diagram which shows the function structure of the information processing apparatus which concerns on 3rd Embodiment. It is a figure which shows an example of the mounting form of the environment specific | specification part which concerns on 3rd Embodiment.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
(First embodiment)
First, an outline of the first embodiment of the present invention will be described. FIG. 1 is a conceptual diagram showing an outline of the present embodiment.
The information processing apparatus 10 (FIG. 3) according to the present embodiment mainly executes two-stage processing. The two-stage processing is (I) learning step and (II) organism identification (organism identification) step. (I) Before executing the process of the learning step, an environmental sample containing the organism is collected in advance from the organism existence site, which is the environment in which the organism exists. Then, a metagenomic analysis is performed on the collected environmental samples. In the metagenomic analysis, the base sequence of DNA (Deoxiribo Nucleic Acid) is obtained from an environmental sample. By metagenomic analysis, a composition based on the relative DNA amount of the organism is obtained based on the DNA sequences of the organisms contained in the environmental sample. In addition to metagenomic analysis, a composition based on the relative RNA amount of an organism can be obtained from an RNA (Ribonucleic Acid) sequence obtained by metatranscriptome analysis. Thus, based on the relative DNA amount and the relative RNA amount, an organism existing in the environment is identified, and an organism having the largest relative amount is defined as a dominant organism. Information indicating at least the identified biota is input to the information processing apparatus 10 as biological information of the living organisms present in the environment. The biological information is, for example, biota information indicating the abundance for each identified organism, or information indicating the dominant organism with the highest abundance ratio. On the other hand, as environment information, an image representing the environment of the living organism part is captured with a camera. As a camera used for imaging, for example, a camera built in various electronic devices such as a mobile phone and a tablet terminal device can be used. At least a captured image is input to the information processing apparatus 10 as environment information related to the environment in which the organism exists. (I) In the learning step, the information processing apparatus 10 performs machine learning using teacher data with environmental information as input and biological information as output, and generates a biological information model as a learning result. The teacher data includes environmental information and biological information. As faculty data, for example, biota data reanalyzed from public base sequence databases published by DDBJ (DNA Data Bank of Japan) and environmental images that match environmental information in public database metadata May be. As the machine learning method, for example, a technique such as deep learning can be used.

  (II) In the organism identification step, a captured image representing the environment to be analyzed is input to the information processing apparatus 10. (I) An image different from the captured image input in the learning step is generally used as an input in the (II) organism identification step. The information processing apparatus 10 specifies a living organism corresponding to the input captured image based on the biological information model generated in the (I) learning step. In the example illustrated in FIG. 1, when the environment of the structure wall surface including biological contamination is input as a captured image, the information processing apparatus 10 specifies cyanobacteria (Hassalia) as a living organism corresponding to the input captured image. The information processing apparatus 10 may output the name of the specified organism or the biota obtained by DNA analysis. Thereby, the organism can be identified from the captured image without collecting the organism from the environment where the organism exists and performing a DNA analysis experiment.

(Computer system configuration)
Next, the configuration of the computer system 100 included in the information processing apparatus 10 according to the present embodiment will be described.
FIG. 2 is a schematic block diagram showing the configuration of the computer system 100.
The computer system 100 includes a communication unit 101, an input unit 102, an output unit 103, a storage unit 104, and a CPU (Central Processing Unit) 105. These components are connected via a bus B so that they can communicate with each other.

The communication unit 101 is a communication module such as a communication IC (Integrated Circuit).
The input unit 102 is, for example, an input interface. The input unit 102 may be an input module including a pointing device such as a mouse and a touch pad, a keyboard, a microphone, and a camera. The input unit 102 may be configured integrally with a display as a touch panel.
The output unit 103 is, for example, an output interface. The output unit 103 may be an output module such as a display panel, a speaker, and a woofer. The input interface and the output interface may be configured integrally.

The storage unit 104 includes, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), an EEPROM (Electrically Erasable Programmable Read-Only Memory), a ROM (Read-Only Memory), a RAM (Random Access Memory), and the like. Module. The storage unit 104 is not limited to a built-in one, but may be an external storage module connected by a digital input / output port such as a USB (Universal Serial Bus).
The CPU 105 executes processing specified by instructions described in various programs stored in the storage unit 104 and controls the operation of each component included in the computer system 100.

(Functional configuration of information processing device)
Next, a functional configuration of the information processing apparatus 10 according to the present embodiment will be described.
FIG. 3 is a schematic block diagram illustrating a functional configuration of the information processing apparatus 10 according to the present embodiment.
The information processing apparatus 10 includes an environment information acquisition unit 11, a biological information acquisition unit 12, an image acquisition unit 13, a biological information output unit 14, a data storage unit 15, and a control unit 16. . The information processing apparatus 10 is configured as a server apparatus that includes these units. The functions of the environmental information acquisition unit 11, the biological information acquisition unit 12, the image acquisition unit 13, the biological information output unit 14, the data storage unit 15, and the control unit 16 are stored in advance in the data storage unit 15 by the CPU 105. This is realized by executing the program. A program for realizing the functions of these units may be implemented as a subroutine included in an API (Application Programming Interface). A subroutine is a program that describes a process that is executed when it is called based on an instruction described in another program. When processing described in a subroutine is executed, information that is part or all of the processing result obtained by the processing is configured as an API function that is returned as a return value to a variable or array specified in the calling source program May be. The functions of these units may be realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit). The environment information acquisition unit 11, the biological information acquisition unit 12, and the image acquisition unit 13 are realized using the communication unit 101, the input unit 102, or both. The biological information output unit 14 is realized using the communication unit 101, the output unit 103, or both. A storage unit 104 is used as the data storage unit 15. The information processing apparatus 10 may be configured as a dedicated apparatus including these units.

  The environment information acquisition unit 11 acquires environment information. The environmental information is information indicating an environment in which a living organism exists. The environment information acquired by the environment information acquisition unit 11 is learning environment information mainly used as teacher data. The environment information includes at least a captured image representing the environment. The captured image is an image obtained by capturing the environment. The environment information acquisition unit 11 receives environment information from, for example, other devices such as a mobile phone, a tablet terminal device, and a digital camera. The environment information may include incidental information related to the environment represented by the captured image. The environment information acquisition unit 11 outputs the acquired environment information to the control unit 16. An example of the incidental information will be described later.

  The biological information acquisition unit 12 acquires biological information. The biological information is information including a biological name of an organism existing in the environment indicated by the environmental information. The name of the organism is the name of the organism identified from the base sequence obtained by performing the metagenomic analysis on the environmental sample collected from the environment. The organisms to be identified are all living organisms including microorganisms that cannot be visually recognized unless enlarged using a microscope. Microorganisms include eukaryotes including viruses, fungi and the like, as well as prokaryotes such as eubacteria (bacteria) and archaea (archia). The class for identifying the name of the organism may be any class such as a world, a gate, a family, a genus, and a species. In the following description, the distinction of an organism obtained by identification may be referred to as an organism type or simply a type. The biological information may be information indicating a biota in the environment. The biota means all kinds of organisms existing in a specific environment, and may be represented by an abundance for each organism name. The abundance may be represented by any of the number of individuals in the sample, the DNA density, the number of reads, or their constituent ratio (eg, relative DNA amount). Further, the biological information may be information indicating the name of the dominant organism that is the name of the dominant organism in the environment. A dominant organism is an organism that has more abundance than other organisms in its environment. In the present embodiment, among the organisms constituting the biota, in particular, the organism having the maximum relative DNA amount. The biological information acquisition unit 12 outputs the acquired biological information to the control unit 16.

Similar to the environment information acquisition unit 11, the image acquisition unit 13 acquires environment information including a captured image. The environmental information acquired by the image acquisition unit 13 is diagnostic environmental information mainly for specifying a living thing. The diagnostic environment information is acquired independently of the learning environment information acquired by the environment information acquisition unit 11. In general, the captured image included in the diagnostic environment information is an image representing a separate environment from the captured image included in the learning environment information. In the following description, the environment information acquired by the image acquisition unit 13 is called diagnostic environment information, and is distinguished from the environment information (learning environment information) acquired by the environment information acquisition unit 11. The image acquisition unit 13 outputs the acquired diagnostic environment information to the control unit 16.
The biological information output unit 14 outputs biological information input from the control unit 16. The biological information output unit 14 may transmit biological information as a response to the device that is the transmission source of the diagnostic environment information.
The data storage unit 15 stores various data used by the control unit 16 and various data generated by the control unit 16.

The control unit 16 controls each component of the information processing apparatus 10. The control unit 16 includes a machine learning model learning unit 161 and an organism identification unit 162.
The learning unit 161 receives environmental information from the environmental information acquisition unit 11 and biological information from the biological information acquisition unit 12. The learning unit 161 associates biological information indicating living organisms existing in the environment as a target output, and associates the information as a set of teacher data having environmental information related to the environment as an input. The learning unit 161 obtains a model indicating a function that maps to a target output as an output corresponding to an input using a predetermined machine learning algorithm (supervised learning). The learning unit 161 uses a plurality of sets of teacher data, and recursively calculates parameters constituting the model so that the difference between the output corresponding to the input and the target output is small. The learning unit 161 performs, for example, deep learning when acquiring a model. Deep learning is machine learning using a multi-layered structure, particularly a neural network having three or more layers. As a neural network having a multilayer structure, for example, a convolution network can be used. A convolutional network is a type of forward-propagating neural network that is not fully connected. The learning unit 161 uses, for example, a gradient descent method in model learning. The learning unit 161 determines that the model has converged when the magnitude of the difference between the output corresponding to the input and the target output is less than a predetermined magnitude, and the model obtained as a learning result is used as data as a biological information model. Store in the storage unit 15. The biological information model is a model indicating a function that maps to biological information as an output corresponding to environmental information.

  The organism identification unit 162 reads the biological information model from the data storage unit 15 and obtains biological information corresponding to the diagnostic environment information input from the image acquisition unit 13 using the function indicated by the read biological information model. The organism identification unit 162 uses a function common to the function employed by the learning unit 161. For example, when the learning unit 161 uses a multilayered neural network, the organism identification unit 162 also uses a multilayered neural network. The organism identification unit 162 outputs the obtained organism information to the organism information output unit 14.

  The organism identification unit 162 may output the biological information corresponding to the diagnostic environment information including a single still image to the biological information output unit 14, or may take an image at a certain time (Time Lapse imaging, interval imaging) in a certain environment. ) May be output to the biological information output unit 14 corresponding to the captured image (time-series image). Thereby, temporal variation of the biota indicating the biota at each imaging time point is obtained. As a time interval for capturing an image, a time interval having a sufficient scale for observing a significant temporal variation of the biota may be set in advance. This time interval generally differs depending on the type of biota / dominant organism of interest.

(Neural network configuration example)
Next, a configuration example of a neural network used in the learning unit 161 and the organism identification unit 162 according to this embodiment will be described.
FIG. 4 is a diagram illustrating a configuration example of the neural network according to the present embodiment.
FIG. 4 shows a neural network of L + 1 (L is an integer of 1 or more) layers. The neural network shown in FIG. 4 includes an input layer, an intermediate layer of an L layer, and an output layer. The input layer includes I (I is an integer of 2 or more) input terminals In-1 to In-I. Input values x _{in_1 to} x _{in_I} are respectively input to the input terminals In-1 to In-I. Input value _x IN_1 _{~x IN_i} is a value which constitutes the input information, respectively. Each input end, and outputs an output value corresponding to each input value to each node _{Md-1,1~Md-1, K} 1 of the first layer of the intermediate layer. The l (l is an integer from 1 to L) _layer, K l _{(K l} is an integer of 2 or more) number of nodes Md-l, 1~Md-l, configured to include a _{K l.} An output value output from each node of the (1-1) th layer is input to each node of the lth layer as an input value. Here, each node of the 0th layer corresponds to an input end. Each node in the l-th layer calculates an output value for the input value, and outputs the calculated output value to the node in the l + 1-th layer. Each node in the (L + 1) th layer corresponds to an output end. The output layer includes M (M is an integer of 2 or more) output terminals Out-1 to Out-M. The output terminal Out-1~Out-M, output values respectively outputted from each node of the L layers is input as an input value _{y in_1 ~y in_M.} The output terminals Out-1 to Out-M calculate output values y _out — ₁ to y _{out_M} corresponding to the input values, and output the calculated output values y _out — ₁ to y _{out_M} . The output values y _{out —} 1 to y _out — _M are values that constitute output information.

  In the learning step, the learning unit 161 calculates an activation function parameter that gives the relationship between the input value and the output value at each of the input end, the node, and the output end. The calculated parameter is a value constituting a model obtained by learning. As the activation function, for example, any of a ramp function, a soft sign, a second-order polynomial, or the like is used. The learning unit 161 uses an error function indicating the magnitude of the difference between the output with respect to the input and the target output in determining convergence. The error function is, for example, one of the following functions. (1) The sum of squares of the difference between the output value constituting the target output and the output value as the output with respect to the input; Both functions are functions indicating that the larger the value obtained, the greater the difference. The organism identification unit 162 calculates output values corresponding to the input values given to the input end, the node, and the output end using the parameters of the activation function constituting the obtained model.

(Example of teacher data)
Next, an example of teacher data will be described. The teacher data according to the present embodiment includes a plurality of sets of environmental information and biological information. FIG. 5 is a diagram illustrating an example of one set of teacher data. The environmental information may include incidental information regarding the environment. The incidental information includes any one or all of geospatial information, surrounding biological information, weather information, building / structure information, and date / time. The geospatial information is information indicating the geographical position of the environment, for example, latitude and longitude depth. The surrounding organism information is information indicating living organisms existing around the environment. The meteorological information is information indicating the climate, temperature, precipitation, relative humidity, sunshine duration, wind direction, wind speed, earth temperature, water temperature, etc. in the environment. The building / structure information is information indicating the positional relationship, the size of the space, the sunshine state, the material, the elapsed year, the surface moisture content, the ventilation state, the direction, and the like in the building or structure that constitutes the environment. The positional relationship includes information such as indoor / outdoor distinction, floor number, part, and the like. The part is, for example, a ceiling, a wall surface, a floor surface, or the like. The date / time indicates the date / time, time zone, season, and the like when the image was captured. The incidental information may be described as metadata of a predetermined image file format. These pieces of information are information that affects the growth of organisms in the environment. Moreover, such information may be provided with the base sequence of a biota as a biota annotation by a base sequence database. As the image file format, for example, Exif (Exchangeable image file format), XPM (Extensible Metadata Platform), or the like may be used. Further, the incidental information may include information on photographing equipment model information that influences image quality, image photographing conditions such as the presence or absence of a flash, and usage conditions of the microscope such as macro lens mounting conditions and magnification conditions.

  When the learning unit 161 and the organism identification unit 162 use a neural network, a pixel value for each pixel indicating an image may be used as an input value to each input terminal. Further, each parameter constituting a feature amount indicating the morphological feature of the image may be used as an input value to each input terminal. As the feature amount, for example, a code value indicating the direction of the ridge for each block obtained by dividing the image, a flag value indicating whether the pixel belongs to the contour of the subject, or the like can be used. In addition, a code value or a flag value indicating element information constituting the incidental information may be used as an input value to each input terminal.

  On the other hand, when the learning unit 161 and the organism identification unit 162 use a neural network, the abundance for each organism constituting the biota indicated by the organism information may be used as an output value from each output terminal. When the biological information includes class information for identifying the organism name as the biota analysis condition, the learning unit 161 and the organism identification unit 162 determine the class by referring to the biota annotation. May be. As biological information, the name of the dominant organism may be used instead of the biota. In this case, the learning unit 161 gives 1 as the predetermined first value as the output value of the target output corresponding to each dominant organism name, and the predetermined value as the output value of the target output corresponding to the other organism names. Model learning may be performed by giving 0 as the second value. The organism identification unit 162 preferentially selects an organism name corresponding to an output value that obtains a positive value higher than a predetermined threshold, which is a difference from the second value, as an output value from the output end with respect to an input value to the input end. It may be specified as an occupying name. Since many kinds of organisms exist in the environment, the number of dimensions of output information, that is, the number of output terminals is generally large. By using the dominant organism name, the degree of freedom related to the identification of the organism from the output value is reduced. In the above-described example, it is sufficient if a significant difference between the output value for each output end and the second value is determined. Therefore, the amount of calculation related to model learning is reduced. In addition, although the case where the output value of each one output end is matched with one kind of living thing was mentioned as an example in the above, it is not restricted to this. The output value of one output terminal may be associated with a plurality of types of organisms. For example, symbiotic organisms that coexist in the same environment may be employed as the plurality of types of organisms.

  The learning unit 161 may perform ensemble learning when generating the biological information model. Ensemble learning is a process in which a plurality of different models are independently learned, and output values based on the models are integrated to obtain a final output value. A part of parameters representing environment information may be used as input values used for learning of each model. However, the set of input values used for learning differs between models, and all the parameters representing environment information are covered by the set of input values for each model. In the integration of output values, a weighted average is made for each output value. The weight coefficient in each model used for the weighted average is also a learning target. The organism identification unit 162 calculates an output value corresponding to the input value based on the model learned for each model, and finally calculates the sum of the multiplication values obtained by multiplying the output value calculated for each model and its weight coefficient. Calculate as output value. Thereby, the biological information based on the final output value can be determined more accurately.

  The organism identification unit 162 can determine an organism present in the environment based on the diagnostic environment information including an image representing the environment. Therefore, unlike the prior art, microorganisms existing in the environment can be identified without enlarging the image with a microscope. Moreover, even if the living thing which exists in an environment cannot be visually recognized, the living organism specific | specification part 162 can identify the living organism. Non-visible organisms are, for example, organisms such as bacteria that cannot be seen as they are because each individual is fine. According to the present embodiment, it is not necessary to cultivate organisms collected from the environment to form a colony that can be visually recognized or to enlarge such an organism using a microscope. In the present embodiment, the distinction between visible and non-visible organisms is not based on biological classification, but whether or not it is visible in the environment, that is, whether or not it is expressed as an image. Based. For example, a colony formed by breeding is treated as a visible organism even if it is a fungus that is an invisible organism alone. Note that the organism identification unit 162 may identify an organism imaged using a magnified microscope image captured by attaching a macro lens to the camera as diagnostic environment information.

(Mounting form of the organism identification part)
Next, an implementation form of the organism identification unit 162 will be described. The implementation form of the organism identification unit 162 includes initial integration (Early Integration) and late integration (Late Integration). The initial integration is an implementation form in which the biological information corresponding to the input diagnostic environment information is directly specified using the biological information model generated by the learning unit 161 as described above (FIG. 6). In order to acquire biological information having reliability, it is necessary to learn a biological information model using a lot of teacher data. Since the set of environmental information and biological information is not sufficiently stored at the beginning of operation, there is a possibility that reliable biological information cannot be obtained.

  Therefore, until the amount of the set of environmental information and biological information used as teacher data reaches a predetermined amount, the biological identification unit 162 may adopt late integration. The organism identification unit 162 may adopt initial integration after the amount of the set of environment information and organism information is equal to or greater than a predetermined amount and the organism information model based on the set is obtained.

  Late integration is an implementation that identifies other information corresponding to the input diagnostic environment information and indirectly identifies biological information corresponding to the identified other information. As other information, information having a dependency relationship with both environmental information and biological information and having a lower degree of freedom than biological information is used. The other information is, for example, location information and visible biological information described later. Such information has a known correspondence with biological information. Therefore, even when the set of environmental information and biological information is not sufficiently accumulated as teacher data, biological information with higher reliability can be obtained.

FIG. 7 is a diagram illustrating an example of late integration, which is an implementation form of the organism identification unit 162.
In the example illustrated in FIG. 7, the organism identification unit 162 obtains location information corresponding to the input diagnostic environment information using a function indicated by the scene / object information model. The organism identification unit 162 determines the biological information corresponding to the identified scene / object information with reference to the scene / object-biota data indicating the scene / object-biota correspondence dictionary. The scene / object information refers to one or both of scene information and object information.

  The scene / object information may be the same as the geospatial information and the building / structure information described above. In the example shown in FIG. 7, information such as bricks, which are surface objects of buildings in the environment, is used as object information, and “outdoor warehouse” is used as scene information. Geospatial information may be further added to the scene / object information. Accordingly, the learning unit 161 inputs environment information and scene / object information, and performs machine learning as a target output to generate a scene / object information model. The learning unit 161 stores the generated scene / object information model in the data storage unit 15 in advance. The data storage unit 15 further stores scene / object-biota data. The scene / object-biota data is data configured by associating bioinformation indicating a biota with each scene / object information. The organism identification unit 162 refers to these data when defining the organism information. The example shown in FIG. 7 can be applied to a non-visible creature that cannot visually recognize a living creature that exists in the environment, or a visually-visible creature that can be visually recognized.

FIG. 8 is a diagram illustrating another example of late integration, which is an implementation form of the organism identification unit 162.
In the example illustrated in FIG. 8, the organism identification unit 162 obtains visual biological information corresponding to the input diagnostic environment information using a function indicated by the visual biological information model. The organism identification unit 162 determines the biological information corresponding to the identified visually-identified biological information with reference to the visually-identified biological-biota data indicating the visually-identified biological-biota correspondence dictionary.

  Visible organism information is information indicating an organism that is mainly visually recognized in an environment where an image is captured. This organism appears in the captured image. In the example illustrated in FIG. 8, hypnum (genus Hyphaceae) is used as the visible biological information (visualized biological name). Therefore, the learning unit 161 inputs environmental information and visible biological information, respectively, and performs machine learning as a target output to generate a visual biological information model. The learning unit 161 stores the generated visually recognized biological information model in the data storage unit 15 in advance. The data storage unit 15 further stores visible bio-biota data. Visible organism-biota data is data obtained by associating biometric information indicating a biota for each visible organism. In the biota, the abundance of each organism other than the visible organism is indicated. These creatures may include creatures that are not visible. The organism identification unit 162 refers to these data when defining the organism information. The example shown in FIG. 8 can be applied regardless of the presence or absence of a non-viewable creature that cannot be seen in an environment in which a visible creature that can be seen exists.

As described above, the information processing apparatus 10 according to the present embodiment includes the learning unit 161, the image acquisition unit 13, and the organism identification unit 162. The learning unit 161 is based on teacher data in which environmental information including at least a captured image obtained by imaging an environment in which a living organism is present is associated with biological information that is a result of identifying a living organism that exists in the environment based on a base sequence. Machine learning. The image acquisition unit 13 acquires a captured image obtained by capturing an environment. The organism identification unit 162 identifies an organism corresponding to the image acquired by the image acquisition unit 13 based on the result of machine learning by the learning unit 161.
According to this configuration, a living thing existing in the environment represented by the captured image acquired by the image acquisition unit 13 is estimated. Therefore, organisms present in the environment are easily identified through genome level annotation information without being collected in the environment. Here, in the identification of organisms, visualization means such as culture and observation with an optical microscope are not required. In addition, it is possible to identify an organism that cannot be visually recognized.

(Second Embodiment)
Next, a second embodiment of the present invention will be described. About the same structure as the above-mentioned embodiment, the same code | symbol is attached | subjected and the description is used. In the following description, differences from the above-described embodiment are mainly used.
FIG. 9 is a schematic block diagram illustrating the configuration of the information processing system 1 according to the present embodiment.
The information processing system 1 according to the present embodiment includes an information processing apparatus 10 and an electronic device 20. The information processing apparatus 10 and the electronic device 20 are connected via a network and can transmit and receive various data.

  The electronic device 20 has a function of acquiring environmental information including a captured image and transmitting the acquired environmental information to the information processing apparatus 10. The electronic device 20 may include a camera for capturing an image. The electronic device 20 may have a function of receiving biological information from the information processing apparatus 10 and presenting the received biological information. For example, the electronic device 20 may be a mobile terminal such as a mobile phone or a tablet terminal device, an inspection device for inspecting the environment, or a control device such as ventilation control in a building.

Next, the configuration of the information processing apparatus 10 will be described. In the following explanation, a case where the present invention is applied to the proposal of a drug according to the presence state of a living organism in the environment and the notification of a change in the presence state with time is taken as an example.
The information processing apparatus 10 includes an environment information acquisition unit 11, a biological information acquisition unit 12, an image acquisition unit 13, a biological information output unit 14, a data storage unit 15, and a control unit 16. . The control unit 16 includes a learning unit 161, a biological identification unit 162, and a biological information processing unit 163. Compared with the information processing apparatus 10 illustrated in FIG. 1, the control unit 16 according to the present embodiment further includes a biological information processing unit 163.

  The organism information processing unit 163 identifies the dominant organism name based on the organism information input from the organism identification unit 162. If the input biological information is information that clearly indicates the dominant organism name, the organism information processing unit 163 adopts the dominant organism name as it is. When the input biological information is information indicating a biota, the biological information processing unit 163 determines the biological names of a predetermined number of living organisms as dominant biological names in descending order from the organism having the largest abundance.

  The biological information processing unit 163 refers to the drug data stored in advance in the data storage unit 15 and identifies the drug corresponding to the same organism name as the identified dominant organism name. The biological information processing unit 163 transmits drug information indicating the specified drug to the electronic device 20 via the biological information output unit 14. The drug information may be transmitted in association with the biological information indicating the dominant organism name. The electronic device 20 presents the drug information received from the information processing apparatus 10.

(Drug data)
Next, an example of drug data is shown. FIG. 10 is a diagram illustrating an example of drug data. In the example shown in FIG. 10, the organism name 1 and the organism name 2 are associated with the drug 1 and the drug 2, respectively. As a chemical | medical agent, you may apply to the control agent aiming at destroying from a structure, a structure (a waterway, an offshore installation etc. are included), a crop field (farmland), and other environments. When the organisms to be destroyed are weeds and fungi, they correspond to herbicides and fungicides, respectively. As the name of the organism recorded in the drug data, the organism name of an organism that may cause biocontamination or biodeterioration risk may be used. Biological contamination refers to a state in which an organism has propagated in an artificial environment such as a crop field, a structure, or a building. Biological degradation risk refers to the possibility or degree of deterioration of structures, buildings, etc. due to the propagation of organisms. Therefore, the electronic device 20 can acquire and present information on the dominant organism name and the controlling agent of the dominant organism in the captured environment based on the diagnostic environment information including the acquired image.
In addition, a chemical | medical agent is not necessarily restricted to a control agent, Depending on the application purpose, the chemical | medical agent etc. which can control the presence or absence and growth of the living organisms, such as a carbon source, a vitamin, and inorganic salts, may be sufficient.

(Changes in the presence of organisms over time)
Next, the case of notifying the time change of the existence state of a living organism will be described.
The electronic device 20 may sequentially transmit diagnostic environment information including captured images obtained by capturing a predetermined environment at regular intervals to the information processing apparatus 10.
The organism identification unit 162 of the information processing apparatus 10 determines biological information corresponding to the environmental information for diagnosis at each time point. The organism information processing unit 163 identifies the dominant organism name at that time based on the organism information sequentially input from the organism identification unit 162. Therefore, the biological information processing unit 163 determines whether there is a change in the dominant biological name specified last time. When the biological information processing unit 163 determines that there is a change, the biological information processing unit 163 causes the biological information output unit 14 to transmit biological information indicating the dominant organism name determined to have the change. The biological information processing unit 163 does not transmit biological information to the biological information output unit 14 when determining that there is no change. Thereby, the electronic device 20 can notify the dominant organism name when the dominant organism name changes, and can avoid notifying the dominant organism name when it does not change. When the biological information processing unit 163 determines that there is a change, the biological information processing unit 163 may cause the biological information output unit 14 to transmit the substance information indicating the substance corresponding to the dominant biological name to the biological information in association with the biological information. Good. At this time, when the name of the dominant organism fluctuates, the electronic device 20 can present information on a substance related to the dominant organism.

  As described above, the organism specifying unit 162 may sequentially specify the biota as the biological information corresponding to the environmental information including the images captured at regular time intervals. In that case, the biological information processing unit 163 may determine whether or not the abundance associated with the dominant organism name identified at each time exceeds a predetermined abundance threshold. When the biological information processing unit 163 determines that the abundance exceeds the threshold (indicated by a broken line in FIG. 11), the biological information output unit 14 transmits the biological information indicating the dominant biological name to the electronic device 20. . When the biological information processing unit 163 determines that the threshold value is not exceeded, the biological information output unit 163 does not cause the biological information output unit 14 to transmit the biological information to the electronic device 20. As a result, the electronic device 20 can notify the user of the dominant organism name when the abundance exceeds a predetermined abundance threshold, and when the abundance does not exceed the prescribed abundance threshold. Do not notify the dominant organism name. When the biological information processing unit 163 determines that the abundance exceeds a predetermined abundance threshold, the biological information output unit 14 associates the substance information indicating the substance corresponding to the dominant organism name with the biological information. However, it may be transmitted to the electronic device 20. Therefore, when the abundance becomes larger than the predetermined abundance, the electronic device 20 prompts the user to use the presented substance by presenting information on the substance related to the dominant organism. it can.

  It should be noted that the organism whose type is to be determined, that is, the output target, may be an organism that is set in advance in the organism identification unit 162. The target organisms are, for example, fungi and bacteria whose attention is paid to the risk of contamination due to breeding. The organism to be set may be an organism whose dominant organism name is specified based on the organism information from the organism specifying unit 162 or may not be the organism. In addition, the abundance to be determined may be the abundance of the organism related to the name of each organism, the abundance of the entire organism related to the dominant organism name, or the abundance of the entire organism to be set. Also good.

(Evaluation of biodegradation risk)
Note that the biological information processing unit 163 may calculate a risk evaluation value that quantitatively indicates a biological degradation risk with respect to the artificial environment, as an index value related to the abundance of the living organism. The organism information processing unit 163 calculates, for example, a sum obtained by multiplying the abundance of organisms for each type of determination target by a weighting factor corresponding to the organism name or a discrete value thereof as a risk evaluation value. The weighting coefficient is a value indicating the degree of contribution to the biological degradation risk that differs for each type of organism. The biological information processing unit 163 transmits the calculated risk evaluation value to the biological information output unit 14. The electronic device 20 can present the risk evaluation value information received from the information processing apparatus 10. Therefore, the user can grasp the biodegradation risk in the environment. In addition, the biological information processing unit 163 may determine whether to transmit the substance information, the dominant organism name, or the risk evaluation value itself depending on whether the risk evaluation value exceeds a predetermined threshold value of the risk evaluation value. Good.

(Acquisition of dominant organism distribution)
Further, the image acquisition unit 13 may receive diagnostic environment information from each of the electronic devices 20 existing at a plurality of geographically dispersed points. The organism identification unit 162 acquires biological information corresponding to the diagnostic environment information acquired at each point. The biological information processing unit 163 aggregates the dominant organisms indicated by the biological information at each point to form a dominant organism distribution. Then, the biological information processing unit 163 transmits dominant biological distribution information indicating the dominant biological distribution formed in the biological information output unit 14 to any of the electronic devices 20. Thereby, the electronic device 20 can present the dominant organism distribution indicated by the dominant organism distribution information received from the information processing apparatus 10. The electronic device 20 may represent the abundance of dominant organisms at each point forming the dominant organism distribution using an isoline map (contour diagram) or the like.

(Prediction of breeding areas of dominant organisms)
In addition, the biological information processing unit 163 acquires a time series of the abundance of a predetermined organism from the biota indicated by the biological information at each point, performs a prediction process based on these time series, and temporally from the present time. The abundance at a later future time may be calculated. The time series of the abundance consists of abundances at a plurality of time points up to the present time. As a prediction processing method, for example, an existing prediction algorithm such as a linear prediction method can be used. Then, the biological information processing unit 163 may aggregate the abundance for each type of organism at a future time point to form a biological distribution at each point. Here, the biological information processing unit 163 selects a dominant organism at one of the processing target points from the formed organism distribution, and indicates the dominant organism distribution indicating the abundance at each point related to the selected dominant organism. Form. Then, the biological information processing unit 163 transmits dominant biological distribution information indicating the dominant biological distribution formed in the biological information output unit 14 to any of the electronic devices 20. In this case, the electronic device 20 can present the dominant biological distribution at the future time point. Therefore, the user can grasp the type of dominant organism and its breeding area at the future time point.

(Modification)
Next, a modification of this embodiment will be described. The information processing system 1 according to the present modification supports the operation of a device that is installed in the environment and controls the state of the environment according to the presence state of the organism in the environment indicated by the diagnostic environment information. In the following description, a case where the device to be controlled is a ventilator (not shown) installed in a room of a house or other building is taken as an example.

The electronic device 20 is an IoT (Internet of Things) device that acquires detection information from various sensors (not shown) installed in a room that is an installation environment of a ventilation device.
In the room, a temperature sensor for detecting the temperature, a first opening / closing sensor for detecting opening / closing of the door, and a second opening / closing sensor for detecting opening / closing of the window are installed. Air temperature information indicating the detected air temperature is input from the temperature sensor to the electronic device 20, and first opening / closing information indicating opening / closing of the detected door is input from the first opening / closing sensor to indicate opening / closing of the detected door. Input from the second open / close sensor.

The electronic device 20 includes a camera that images a predetermined part of the room, for example, a wall surface. The camera captures images every predetermined time (for example, 5 to 30 minutes). The electronic device 20 integrates the first opening / closing information, the second opening / closing information, and the temperature information at each time point when the camera captures an image to form detection information. The electronic device 20 transmits the diagnostic environment information including the captured image captured at that time and the detection information to the information processing system 1 in association with each other.
The electronic device 20 receives control information as a response to the transmitted diagnostic environment information and detection information. The electronic device 20 controls the output of the ventilation device based on the control information received from the information processing device 10. The electronic device 20 may be configured integrally with the ventilation device.

  The biological information processing unit 163 of the information processing apparatus 10 receives biological information corresponding to the diagnostic environmental information at each time point from the biological identification unit 162 and detection information transmitted along with the diagnostic environmental information. The biological information processing unit 163 determines control information based on the input detection information and biological information. The biological information processing unit 163 specifies control information corresponding to the detection information and the biological information with reference to control data stored in advance in the data storage unit 15, for example. The biological information processing unit 163 transmits the determined control information to the electronic device 20 via the biological information output unit 14. In this modification, the biological information output unit 14 does not necessarily have to transmit biological information.

(Control data)
Next, an example of control data according to this modification will be described. FIG. 12 is a diagram illustrating an example of control data according to the present modification. In the example illustrated in FIG. 12, the control data is data indicating control amounts corresponding to the open / closed state of the door, the open / closed state of the window, the temperature, and the abundance of living things. The value of the air temperature shown in FIG. 12 is a value obtained by discretizing the air temperature into an integer value of 0 to 2. Higher values indicate higher temperatures. The value of the abundance is a value obtained by discretizing the abundance of a predetermined organism into an integer value of 0 to 2. Larger values indicate more abundance. The value of the controlled variable is a value obtained by discretizing the output of the ventilator into any integer value from 0 to 3. A control amount value of 0 indicates that the ventilator is not operated (OFF). A value of the control amount of 1 or more indicates that the ventilator is operated (ON). It shows that the output of a ventilator is so large that this value is large.

  FIG. 12A shows a part that is referred to when the doors and windows indicated by the detection information are both opened and closed. FIG. 12B shows a portion that is referred to when the open / closed states of the door and the window indicated by the detection information are open, closed, or closed and open, respectively. FIG. 12C shows a part that is referred to when the doors and windows indicated by the detection information are both closed. According to this control data, the control amount is smaller as the open state of the room by the door and the window is remarkable, the control amount is larger as the temperature is higher, and the control amount is larger as the amount of the predetermined living being is larger.

Here, the open / closed states of the door and the window indicated by the detection information are open and closed, respectively, the temperature indicated by the detection information is an air temperature within a range corresponding to 2, and the detection amount of the predetermined organism indicated by the biological information Is a detection amount within a range corresponding to 1. In this case, the biological information processing unit 163 determines the value of the control amount as 2 with reference to the portion shown in FIG. The ventilator operates so as to obtain an output having a magnitude specified by 2 which is the value of the determined control amount.
In addition, the open / closed states of the door and window indicated by the detection information are open and open, respectively, the air temperature indicated by the detection information is within the range corresponding to 1, and the detection amount of the predetermined organism indicated by the biological information is Assume that the detection amount is within a range corresponding to 1. In this case, the biological information processing unit 163 refers to the part shown in FIG. 12A in the control data and determines the value of the control amount as 0. Therefore, the ventilator stops its operation. Therefore, the output of the ventilator is controlled so that the indoor open state is significantly smaller, the higher the temperature is, the larger the output is, and the larger the predetermined amount of living organisms is.

As described above, the information processing apparatus 10 according to this embodiment refers to the substance data (for example, drug data) indicating the substance associated with each living thing, and determines the living thing specified by the living organism specifying unit 162. A biological information processing unit 163 that identifies a corresponding substance is provided.
With this configuration, a substance corresponding to a living organism existing in the environment indicated by the captured image acquired by the image acquisition unit 13 is specified. Therefore, the user is encouraged to preserve the environment by using the specified substance. For example, when the substance is a control agent for a predetermined organism, elimination or reduction of the biological contamination due to the extermination of the organism is promoted.

In the information processing apparatus 10 according to the present embodiment, the organism identification unit 162 identifies an organism corresponding to the captured image acquired by the image acquisition unit 13 every predetermined time. The information processing apparatus 10 includes a biological information processing unit 163 that causes the biological information output unit 14 to output biological information indicating the organism specified by the organism specifying unit 162 when detecting a change in the organism specified by the organism specifying unit 162.
With this configuration, when a change in an organism present in the environment indicated by the captured image acquired by the image acquisition unit 13 every predetermined time is detected, biological information indicating the organism is output. When the organism specified from the captured image changes, the user is notified of the biological information indicating the organism, so that the change of the specified organism can be noticed. Further, when the specified organism does not change, the biological information is not sequentially notified to the user by omitting the processing related to the output of the biological information. Therefore, the processing amount for notification is reduced, and the biological information to be notified can be noticed by the user.

In the information processing apparatus 10 according to the present embodiment, the organism identification unit 162 identifies an organism corresponding to the captured image acquired by the image acquisition unit 13 every predetermined time. The information processing apparatus 10 causes the biological information output unit 14 to output biological information indicating the organism identified by the organism identifying unit 162 when the abundance of the organism identified by the organism identifying unit 162 exceeds a predetermined abundance. Part 163.
With this configuration, when the abundance of an organism present in the environment indicated by the captured image acquired by the image acquisition unit 13 every predetermined time exceeds a predetermined abundance, organism information indicating the organism is output. Since the biological information indicating the living organism is notified to the user when there is a large amount of the living organism specified from the captured image, the user can be made aware of the presence of the living organism and the existing amount. Therefore, the user is prompted to take measures against the specified organism. For example, when the specified organism is an organism harmful to the environment, a countermeasure is promoted to reduce or eliminate the biological degradation risk.

In addition, the information processing apparatus 10 according to the present embodiment is based on detection information indicating the state of the installation environment that is an environment indicated by the captured image acquired by the image acquisition unit 13 and the abundance of the organism specified by the organism specifying unit 162. And a biological information processing unit 163 that determines a control amount of a device installed in the installation environment.
With this configuration, the operation of devices installed in the environment is controlled based on detection information indicating the state of the environment indicated by the captured image acquired by the image acquisition unit 13 and the abundance of organisms specified from the captured image. Is done. Therefore, factors that contribute to the abundance of organisms in control are further considered. For example, when the device is a device that controls the environment, such as a ventilator, the abundance of a living organism can be controlled based on the determined control amount.

(Third embodiment)
Next, a third embodiment of the present invention will be described. About the same structure as the above-mentioned embodiment, the same code | symbol is attached | subjected and the description is used. In the following description, differences from the above-described embodiment are mainly used.

FIG. 13 is a schematic block diagram illustrating a functional configuration of the information processing apparatus 10 according to the present embodiment.
The information processing apparatus 10 includes an environmental information acquisition unit 11, a biological information acquisition unit 12, a data storage unit 15, a control unit 16, a biological identification information acquisition unit 17, and an environmental information output unit 18. Is done. The information processing apparatus 10 is configured as a server apparatus that includes these units.
Compared with the information processing apparatus 10 shown in FIG. 1, in the information processing apparatus 10 according to the present embodiment, the image acquisition unit 13 and the biological information output unit 14 are omitted, and the biological identification information acquisition unit 17 and the environment information output unit 18 Furthermore, it is provided. The control unit 16 includes a learning unit 161 and an environment specifying unit 164. In the control unit 16, the organism specifying unit 162 is omitted, and an environment specifying unit 164 is further provided. The biological identification information acquisition unit 17 is realized using the communication unit 101, the input unit 102, or both. The environment information output unit 18 is realized using the communication unit 101, the output unit 103, or both.

  The environment identification unit 164 reads the environment information model from the data storage unit 15 and uses the function indicated by the read environment information model to correspond to the environment corresponding to the organism identification information (biological information) input from the organism identification information acquisition unit 17. Information is obtained (see FIG. 14). The environment information may include information including an image of the environment where the organism indicated by the biota is present. The environment specifying unit 164 uses a function common to the function used for learning the environment information model. The environment specifying unit 164 outputs the obtained environment information to the environment information output unit 18.

  Therefore, the learning unit 161 uses the biological information input from the biological information acquisition unit 12 as a target output, and is input from the environmental information acquisition unit 11 as a set of teacher data that receives the environmental information of the environment where the organism exists. Associate. The learning unit 161 performs supervised learning and obtains a model indicating a function that maps to environment information as an output corresponding to biological information as an environment information model. The learning unit 161 stores the obtained environment information model in the data storage unit 15 in advance.

In addition, when the learning unit 161 and the environment specifying unit 164 use a neural network, the abundance of each biota indicated by the biometric identification information may be used as an input value to each input terminal. When the organism identification information indicates the dominant organism name, the input value to the input terminal corresponding to the dominant organism name may be set to 1, and the input values to the other input terminals may be set to 0.
Further, the output value from each output terminal may be used as the pixel value for each pixel. The output value from each output end may be used as each parameter of the image feature amount. In the present embodiment, the environment specifying unit 164 specifies an image whose pixel value or feature amount is closest to an output value corresponding to the biological identification information from among a plurality of images stored in advance in the data storage unit 15. May be. As the image stored in advance, a part or all of the image included in the teacher data used for learning the environment information model may be used. The above-described error function can be used as an index indicating the degree of approximation, that is, the magnitude of the difference between the pixel value or feature value and the output value. Thereby, even when the number of output terminals is less than the number of pixels or the number of feature parameters, an image representing an environment in which a living thing exists is determined.

The biological identification information acquisition unit 17 receives information similar to the biological information described above as biological identification information, for example, information indicating the biota or dominant organism name in a certain environment. The organism identification information acquisition unit 17 outputs the input organism identification information to the control unit 16.
The environment information output unit 18 outputs the biological information input from the control unit 16. The environment information output unit 18 may transmit the biological information as a response to the device that is the transmission source of the biological identification information.

As described above, the information processing apparatus 10 according to the present embodiment is the result of identifying environmental information including at least a captured image obtained by capturing an environment in which living organisms exist and the living organisms existing in the environment based on the base sequence. A learning unit 161 that performs machine learning based on teacher data in which certain biological information is associated is provided. The information processing apparatus 10 includes an organism identification information acquisition unit 17 that acquires organism identification information, and an environment corresponding to the organism identification information acquired by the organism identification information acquisition unit 17 based on the result of machine learning by the learning unit 161. Is provided with an environment specifying unit 164.
With this configuration, environment information including an image representing an environment in which the organism exists is estimated corresponding to the organism identification information acquired by the organism identification information acquisition unit 17. Accordingly, environmental information indicating an environment in which an organism identified at the genome level exists is provided.

  As described above, the embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to the above, and various design changes and the like can be made without departing from the scope of the present invention. It is possible to

(Appendix)
From the above description, a plurality of aspects of the present invention are grasped as follows, for example. For ease of understanding, reference numerals in the accompanying drawings are added, but the embodiments of the present invention are not limited to the illustrated embodiments.

(Appendix 1) One embodiment of the present invention includes environmental information including at least a captured image obtained by imaging an environment in which an organism exists, and biological information that is a result of identifying an organism present in the environment based on a base sequence. A learning unit 161 that performs machine learning based on the associated teacher data, an image acquisition unit 13 that acquires a captured image obtained by capturing an environment, and the image acquisition unit 13 based on a result of machine learning by the learning unit 161. The information processing apparatus 10 includes an organism identification unit 162 that identifies an organism corresponding to the acquired image.

(Supplementary Note 2) One aspect of the present invention is the information processing apparatus 10 according to Supplementary Note 2, wherein the organism specifying unit 162 refers to substance data indicating a substance associated with each organism. And a biological information processing unit 163 that identifies a substance corresponding to.

(Supplementary Note 3) One aspect of the present invention is the information processing apparatus 10 according to Supplementary Note 1 or Supplementary Note 2, wherein the organism identification unit 162 corresponds to a captured image acquired by the image acquisition unit 13 every predetermined time. A biological information processing unit 163 that outputs biological information indicating the organism specified by the organism specifying unit 162 to the biological information output unit 14 when the organism specifying unit 162 is detected and a change in the organism specified by the organism specifying unit 162 is detected. Prepare.

(Supplementary Note 4) One aspect of the present invention is the information processing apparatus 10 according to any one of Supplementary Note 1 to Supplementary Note 3, wherein the organism identification unit 162 captures images acquired by the image acquisition unit 13 every predetermined time. When the organism corresponding to the image is identified and the abundance of the organism identified by the organism identification unit 162 exceeds a predetermined abundance, the organism information indicating the organism identified by the organism identification unit 162 is sent to the organism information output unit 14. A biological information processing unit 163 for outputting is provided.

(Supplementary Note 5) One aspect of the present invention is the information processing apparatus 10 according to any one of Supplementary Note 1 to Supplementary Note 4, wherein a state of an installation environment that is an environment indicated by a captured image acquired by the image acquisition unit 13 is described. A biological information processing unit 163 that determines a control amount of a device installed in the installation environment based on detection information to be shown and an abundance of the organism specified by the organism specifying unit 162 is provided.

(Appendix 6) One aspect of the present invention is environmental information including at least a captured image obtained by imaging an environment in which a living organism exists, and biological information that is a result of identifying an organism existing in the environment based on a base sequence of a gene; , The learning unit 161 that performs machine learning based on the teacher data, the biological identification information acquisition unit 17 that acquires biological identification information, and the biological identification based on the result of machine learning by the learning unit 161 The information processing apparatus 10 includes an environment specifying unit 164 that specifies an environment corresponding to the organism identification information acquired by the information acquisition unit 17.

(Additional remark 7) One aspect of this invention is the information processing method in the information processing apparatus 10, Comprising: The environmental information which contains at least the picked-up image which imaged the environment where a living thing exists, and the living thing which exists in the said environment on a base sequence A learning step for performing machine learning based on teacher data in association with biological information as a result of identification based on an image, an image acquisition step for acquiring a captured image obtained by imaging the environment, and a result of machine learning by the learning step And an organism identification step for identifying an organism corresponding to the captured image acquired by the image acquisition step.

(Supplementary Note 8) One aspect of the present invention is an information processing method in the information processing apparatus 10, and includes environmental information including at least a captured image obtained by capturing an environment in which a living thing exists, and a living thing existing in the environment as a base sequence A learning step for performing machine learning based on teacher data in association with biological information as a result of identification based on the biological information, a biological identification information acquisition step for acquiring identification information of the biological organism, and a result of machine learning by the learning step And an environment specifying step for specifying an environment corresponding to the organism identification information acquired by the organism identification information acquisition step.

(Supplementary Note 9) One embodiment of the present invention is environmental information including at least a captured image obtained by capturing an environment in which a living organism is captured in a computer, and biological information that is a result of identifying the living organism in the environment based on a base sequence And a learning procedure for performing machine learning based on the teacher data in association with each other, an image acquisition procedure for acquiring a captured image obtained by capturing an environment, and a result of machine learning by the learning procedure. This is a program for executing an organism identification procedure for identifying an organism corresponding to an acquired captured image.

(Supplementary Note 10) One embodiment of the present invention is an environment information including at least a captured image obtained by imaging an environment in which a living organism is present in a computer, and biological information that is a result of identifying an organism existing in the environment based on a base sequence. Learning procedure for performing machine learning based on teacher data in association with each other, a biological identification information acquisition procedure for acquiring biological identification information, and acquisition of the biological identification information based on a result of machine learning by the learning procedure And an environment specifying procedure for specifying an environment corresponding to the organism identification information acquired by the procedure.

DESCRIPTION OF SYMBOLS 1 ... Information processing system, 10 ... Information processing apparatus, 11 ... Environmental information acquisition part, 12 ... Biological information acquisition part, 13 ... Image acquisition part, 14 ... Biological information output part, 15 ... Data storage part, 16 ... Control part, DESCRIPTION OF SYMBOLS 17 ... Biological identification information acquisition part, 18 ... Environmental information output part, 20 ... Electronic device, 100 ... Computer system, 101 ... Communication part, 102 ... Input part, 103 ... Output part, 104 ... Memory | storage part, 105 ... CPU, 161 ... Learning unit, 162 ... Biological identification unit, 163 ... Biological information processing unit, 164 ... Environment identification unit

Claims (11)

Machine learning based on teacher data in which environmental information including at least captured images obtained by imaging an environment in which living organisms are present and biological information that is a result of identifying living organisms existing in the environment based on base sequences are associated with each other. Learning department to perform,
An image acquisition unit that acquires a captured image of the environment;
Based on the result of machine learning by the learning unit, a living organism specifying unit that specifies a living organism corresponding to the captured image acquired by the image acquisition unit;
An information processing apparatus comprising: Referring to the data showing the substances associated with each organism,
The information processing apparatus according to claim 1, further comprising a biological information processing unit that specifies a substance corresponding to the living organism specified by the biological identification unit. The organism identification unit identifies an organism corresponding to the captured image acquired by the image acquisition unit every predetermined time,
The biological information processing part which outputs the biological information which shows the living thing which the said biological identification part specified to the biological information output part, when detecting the change of the biological organism which the said biological identification part specified. The biological information processing part is provided. Information processing device. The organism identification unit identifies an organism corresponding to the captured image acquired by the image acquisition unit every predetermined time,
The biological information processing part which outputs the biological information which shows the living thing which the said living body specific part specified to the living body information output part when the abundance of the living body which the said living body specific part specified exceeds predetermined | prescribed abundance. The information processing apparatus according to claim 3. Based on the detection information indicating the state of the installation environment, which is the environment indicated by the captured image acquired by the image acquisition unit, and the amount of organisms identified by the organism identification unit, the control amount of the device installed in the installation environment The information processing apparatus according to any one of claims 1 to 4, further comprising a biological information processing unit that determines Machine learning based on teacher data in which environmental information including at least captured images obtained by imaging an environment in which living organisms are present and biological information that is a result of identifying living organisms existing in the environment based on base sequences are associated with each other. Learning department to perform,
A biological identification information acquisition unit for acquiring biological identification information;
Based on the result of machine learning by the learning unit, an environment specifying unit for specifying an environment corresponding to the organism identification information acquired by the organism identification information acquisition unit;
An information processing apparatus comprising: An information processing method in an information processing apparatus,
Machine learning based on teacher data in which environmental information including at least captured images obtained by imaging an environment in which living organisms are present and biological information that is a result of identifying living organisms existing in the environment based on base sequences are associated with each other. Learning steps to perform,
An image acquisition step of acquiring a captured image of the environment;
Based on the result of machine learning by the learning step, a biological identification step for identifying a biological object corresponding to the captured image acquired by the image acquisition step;
An information processing method comprising: An information processing method in an information processing apparatus,
Machine learning based on teacher data in which environmental information including at least captured images obtained by imaging an environment in which living organisms are present and biological information that is a result of identifying living organisms existing in the environment based on base sequences are associated with each other. Learning steps to perform,
A biological identification information acquisition step of acquiring biological identification information;
Based on the result of machine learning by the learning step, an environment specifying step for specifying an environment corresponding to the organism identification information acquired by the organism identification information acquisition step;
An information processing method comprising: On the computer,
A machine based on teacher data in which environmental information including at least a captured image obtained by imaging an environment in which a living organism is present and biological information that is a result of identifying the living organism in the environment based on a base sequence of a gene are associated with each other. A learning procedure for learning,
An image acquisition procedure for acquiring a captured image of the environment;
Based on a result of machine learning by the learning procedure, a living organism identification procedure for identifying a living organism corresponding to the captured image acquired by the image acquisition procedure;
A program for running On the computer,
Machine learning based on teacher data in which environmental information including at least captured images obtained by imaging an environment in which living organisms are present and biological information that is a result of identifying living organisms existing in the environment based on base sequences are associated with each other. Learning steps to perform,
An organism identification information acquisition procedure for acquiring organism identification information;
Based on the result of machine learning by the learning procedure, an environment specifying procedure for specifying an environment corresponding to the organism identification information acquired by the organism identification information acquisition procedure;
A program for running   The program according to claim 9 or 10, which is implemented as a subroutine included in an application programming interface.