KR20240053799A - System for Growing Plant Based on Artificial Intelligence - Google Patents

Abstract

An AI-based plant cultivation system is launched.
This embodiment is a mentor cultivator with the same protocol and operating system. Using a Menti cultivator and a mass production cultivator, optimal growth data is generated by using one Mentor cultivator, an expensive expert plant cultivator for precise analysis, and multiple Menti cultivators, a low-cost plant cultivator, and then using a mass production cultivator to simultaneously measure the size of plant cultivation. Provides an AI-based plant cultivation system that can increase N times.

Description

AI-based plant cultivation system {System for Growing Plant Based on Artificial Intelligence}

One embodiment of the present invention relates to an AI-based plant cultivation system.

The content described below simply provides background information related to this embodiment and does not constitute prior art.

Recently, there is a trend to grow plants such as lettuce and strawberries indoors not only for eating but also for interior decoration and indoor air purification.

Because each plant is affected by environmental variables such as sunlight, humidity, amount of water, and temperature, plants planted indoors are likely to die as they are not supplied with sufficient light, water, and nutrients. Busy modern people frequently experience this condition. There is a problem that management is cumbersome.

The development of plant cultivation devices that can cultivate plants regardless of season or location while artificially controlling environmental conditions such as light, temperature and humidity, carbon dioxide concentration, and nutrients is being developed.

Plant cultivation equipment allows you to obtain pollution-free, fresh organic vegetables throughout the four seasons regardless of the season and location in a situation where extreme weather events frequently occur, such as recent frequent occurrences of food stability and price surges, and efficient mass production even in a small space. This may be possible.

This embodiment is a mentor cultivator with the same protocol and operating system. Using a Menti cultivator and a mass production cultivator, optimal growth data is generated by using one Mentor cultivator, an expensive expert plant cultivator for precise analysis, and multiple Menti cultivators, a low-cost plant cultivator, and then using a mass production cultivator to simultaneously measure the size of plant cultivation. The purpose is to provide an AI-based plant cultivation system that can increase N times.

According to one aspect of this embodiment, a mentee cultivator that cultivates plants according to a plurality of different first growth environment input values using a plurality of cultivators and generates first cultivation data for the plants; Among the first cultivation data, the growth environment input value closest to the cultivation purpose is selected, the plant is directly cultivated based on the growth environment input value closest to the cultivation purpose, and the cultivation data for the directly cultivated plant is analyzed. Based on the results, a plurality of second growth environment input values are generated, the plurality of second growth environment input values are transmitted back to the mentee cultivator, and then the Nth corresponding to the second growth environment input value is received from the mentee cultivator. Receive cultivation data, select the growth environment input value closest to the cultivation purpose among the Nth cultivation data, directly cultivate the plant based on the growth environment input value closest to the cultivation purpose, and A mentor grower that generates optimal growth environment input values by repeating the process of analyzing cultivation data; It provides a plant cultivation system comprising: a mass production cultivator that mass-produces the plants by receiving the optimal growth environment input value from the mentor cultivator using a plurality of cultivators.

According to this embodiment as described above, a mentor cultivator having the same protocol and operating system. Using a Menti cultivator and a mass production cultivator, optimal growth data is generated by using one Mentor cultivator, an expensive expert plant cultivator for precise analysis, and multiple Menti cultivators, a low-cost plant cultivator, and then using a mass production cultivator to simultaneously measure the size of plant cultivation. It has the effect of increasing N times.

1 is a diagram schematically showing an AI-based plant cultivation system according to this embodiment.
Figure 2 is a diagram schematically showing the internal modules of the mentee cultivator according to this embodiment.
Figure 3 is a flowchart for explaining the AI-based plant cultivation method according to this embodiment.
Figure 4 is a diagram for explaining a growth section according to this embodiment.
Figure 5 is a diagram for explaining the measurement of chlorophyll fluorescence of plants according to this embodiment.
Figure 6 is a diagram showing an RGB image according to this embodiment.
Figure 7 is a diagram showing 3D scanning according to this embodiment.
Figure 8 is a diagram showing an example of chlorophyll fluorescence measurement according to this embodiment.

Hereinafter, this embodiment will be described in detail with reference to the attached drawings.

1 is a diagram schematically showing an AI-based plant cultivation system according to this embodiment.

The AI-based plant cultivation system 100 according to this embodiment includes a mentor cultivator 110, a mentee cultivator 120, and a mass production cultivator 130. Components included in the plant cultivation system 100 are not necessarily limited thereto.

The mentor cultivator 110 refers to an expensive expert plant cultivator for precise analysis. The mentor cultivator 110 refers to a plant cultivator equipped with expensive sensors and cameras (RGB camera, depth camera, IR camera).

The mentor cultivator 110 communicates with the mentee cultivator 120 and receives the external pattern (outward appearance) of plants according to the growth environment input value (cultivation recipe). The mentor cultivator 110 uses machine learning to group the external patterns (outward appearances) of plants. The mentor cultivator 110 extracts samples from machine learning groups and then grows plants using the growth environment input values (cultivation recipes) corresponding to each sample.

The mentor cultivator 110 cultivates plants by performing light quantity control, cooling and heating control, CO2 control, FAN control, and nutrient solution supply control according to the growth environment input value (cultivation recipe) corresponding to each sample.

The mentor cultivator 110 detects secondary metabolites (stress components of plants) for each cultivated plant, and detects the specific secondary metabolites (stress components of plants) in the most amount for each growth section (planting, germination, or harvest). Decide on a group. The mentor cultivator 110 determines a growth environment input value (cultivation recipe) that allows detection of a large number of secondary metabolites (plant stress components) for the determined specific group.

Plants produce primary metabolites in normal growth environments, and the characteristics of the plant are determined by the primary metabolites. When plants are stressed by their natural growth environment, they produce secondary metabolites that are not produced in the artificial growth environment.

In other words, when plants undergo any form of stress, they produce secondary metabolites. For example, strawberries generally contain primary metabolites that give them a strawberry flavor, and when stressed, secondary metabolites (sulfur oxidation enzymes) are produced. Ginseng and wild ginseng have the same seeds, but ginseng grown in an artificial environment is larger than wild ginseng, but contains fewer ingredients than wild ginseng. Wild ginseng grown under stress in a natural environment is much smaller than ginseng, but its components (including secondary metabolites) are higher than those of ginseng.

If a plant is overly stressed, it will wither and die. Therefore, in a growing environment where the plant is stressed without dying, the plant produces secondary metabolites at the maximum level.

The mentor cultivator 110 applies a growth environment that allows plants to receive an appropriate amount of stress and produce secondary metabolites. For example, the mentor grower 110 controls LEDs to stress plants with UV-A and UV-B. The mentor cultivator 110 exerts stress on the plants by controlling the amount of CO2 supplied. The mentor cultivator 110 dissolves specific components that can cause specific stress in the nutrient solution and supplies them to the nutrient solution feeder. The mentor cultivator 110 numbers or quantifies how much stress was applied to each growth environment.

The mentor cultivator 110 calculates the total area of the plant's leaves using cameras (RGB camera, depth camera, IR camera) equipped to detect secondary metabolites (plant stress components) based on directly grown plants. Or calculate the area of the entire shape. The mentor cultivator 110 compares the total area of the leaf or the area of the entire shape with a preset reference value (area value in the absence of stress) and determines the content of secondary metabolites (stress components of the plant) based on directly grown plants. detect. The mentor grower 110 extracts RGB values for the color of the leaves. The mentor cultivator 110 compares the RGB value of the leaf color with a preset reference value (RGB value in the absence of stress) and detects the content of secondary metabolites (stress components of the plant) based on directly grown plants. .

The mentee cultivator 120 refers to a plant cultivator that is relatively cheaper than the mentor cultivator 110. The mentee cultivation unit 120 includes a plurality of cultivation units. Each of the plurality of cultivators included in the mentee cultivator 120 refers to a cultivator capable of cultivating various plants in large quantities using respective growth environment input values (cultivation recipes).

The mass production cultivator 130 includes a plurality of cultivators. Each of the plurality of cultivators included in the mass production cultivator 130 refers to a cultivator for mass producing one specific plant.

The mass production cultivator 130 controls light intensity, cooling and heating control, CO2 control, FAN control, and nutrient solution supply control according to the growth environment input value (cultivation recipe) for each growth section included in the entire cycle growth environment input value (full cycle cultivation recipe). mass production of one type of specific plant.

The mentor cultivator 110, the mentee cultivator 120, and the mass production cultivator 130 communicate with each other using the same protocol and the same operating system.

The growth and cultivation system according to this embodiment separates and manages the period from sowing and germination of plants to harvesting into preset time series units.

The mentee cultivator 120 creates a plurality of cultivation environments and grows plants using a plurality of different growth environment input values (cultivation recipes) for each growth section (seeding, germination, or harvest). The mentee cultivator 120 extracts the external pattern (outward appearance) of plants grown using a plurality of growth environment input values (cultivation recipes). The mentee cultivator 120 transmits the external pattern (outward appearance) of the plants grown due to the growth environment input value (cultivation recipe) to the mentor cultivator 110.

The mentor cultivator 110 is a growth environment input value (cultivation recipe) transmitted by each of the plurality of cultivators included in the mentee cultivator 120, and a growth environment input value (cultivation recipe) matched for each growth environment input value (cultivation recipe). Machine learning is performed based on the external pattern (outward appearance) of the plants grown as a result, and grouping is performed by external pattern (outward appearance).

The mentor cultivator 110 performs grouping according to the external pattern (outward appearance) grown according to the growth environment input value (cultivation recipe). For example, if there are 100 cultivators (machines) included in the mentee cultivator 120, the mentor cultivator 110 divides 100 external patterns (outward appearance) into groups with similar shapes (e.g., approximately 4 groups (A) group, B group, C group, D group)).

The mentor grower 110 selects one or a preset number of samples from each group to determine which of each group is the most useful. The mentor cultivator 110 extracts growth environment input values (cultivation recipe) for each sample. The mentor cultivator 110 directly cultivates each plant using the growth environment input value (cultivation recipe) for each sample.

The mentor cultivator 110 detects secondary metabolites (stress components of the plant) based on the results of directly cultivating each plant using the growth environment input value (cultivation recipe) for each sample. Here, the mentor cultivator 110 can use an external component analysis server to detect secondary metabolites (stress components of plants).

The mentor cultivator 110 determines a specific group in which the most secondary metabolites (plant stress components) are detected for each growth section (planting, germination, or harvest). The mentor cultivator 110 determines one or more growth environment input values (cultivation recipes) that allow detection of a large number of secondary metabolites (plant stress components) for the determined specific group. The mentor cultivator 110 transmits one or more growth environment input values (cultivation recipes) that produce the most secondary metabolites (plant stress components) to the mentee cultivator 120.

The mentee cultivator 120 receives one or more growth environment input values (cultivation recipes) that produce the most secondary metabolites (plant stress components) from the mentor cultivator 110.

The mentee cultivation unit 120 changes the growth environment of all the cultivation units it contains by reflecting the growth environment input value (cultivation recipe) that generates the most secondary metabolites (stress components of plants).

The mentee cultivator 120 extracts the external pattern (outward appearance) of plants grown with the growth environment input value (cultivation recipe) that generates the most altered secondary metabolites (stress components of plants).

The mentee cultivator 120 displays the external pattern (outward appearance) of plants grown due to the growth environment input value (cultivation recipe) that produces the most changed secondary metabolites (stress components of plants) to the mentor cultivator 110. send to

The mentor cultivator 110 is a plant grown due to the growth environment input value (cultivation recipe) that generates the largest amount of secondary metabolites (plant stress components) transmitted by each of the plurality of cultivators included in the mentee cultivator 120. Machine learning is performed based on the external pattern (outward appearance) of the objects and grouping is performed by external pattern (outward appearance).

The mentor cultivator 110 performs grouping by growth pattern (outward appearance) according to the growth environment input (cultivation recipe) that generates the most secondary metabolites (stress components of plants).

The mentor grower 110 selects one or a preset number of samples from each group to determine which of each group is the most useful. The mentor cultivator 110 extracts growth environment input values (cultivation recipe) for each sample. The mentor cultivator 110 directly cultivates each plant using the growth environment input value (cultivation recipe) for each sample.

The mentor cultivator 110 detects secondary metabolites (stress components of the plant) based on the results of directly cultivating each plant using the growth environment input value (cultivation recipe) for each sample. Here, the mentor cultivator 110 can use an external component analysis server to detect secondary metabolites (stress components of plants).

The mentor cultivator 110 determines a specific group in which the most secondary metabolites (plant stress components) are detected for each growth section (planting, germination, or harvest).

The mentor cultivator 110 determines one or more growth environment input values (cultivation recipes) that allow detection of a large number of secondary metabolites (plant stress components) for the determined specific group.

The mentor cultivator 110 determines one or more growth environment input values (cultivation recipe) that allow detection of a large number of secondary metabolites (plant stress components) for one stage (seeding) of the growth section, and then performs the next stage (germination). Only differential distribution with the previous growth environment input value (cultivation recipe) is detected.

For each growth section, the mentor cultivator 110 transmits one or more growth environment input values (cultivation recipes) that produce the most secondary metabolites (plant stress components) to the mentee cultivator 120. .

The mentor cultivator 110 classifies the period into each stage from the sowing and ignition of a specific plant to the final harvest time for each growth section, and then applies the growth environment input value (cultivation recipe) optimized for each period to the entire growth section. Create a standard corresponding full-cycle growth environment input value (full-cycle cultivation recipe). Once the full cycle growth environment input value (full cycle cultivation recipe) is generated, the mentor cultivator 110 transmits it to the mass production cultivator 130.

The mass production cultivator 130 cultivates a specific plant by applying the growth environment input value (cultivation recipe) for each growth section included in the overall cycle growth environment input value (full cycle cultivation recipe).

The mass production cultivator 130 mass-produces plants by applying growth environment input values (cultivation recipes) divided into each growth section and applying different growth environment input values (cultivation recipes) as the growth section changes.

The mass production cultivator 130 can cultivate a specific plant using the growth environment input value (cultivation recipe) with the lowest statistical error. The mass production cultivator 130 can cultivate a specific plant to produce the highest amount of secondary metabolites (plant stress components).

When the mass production cultivator 130 cultivates a specific plant by applying the growth environment input value (cultivation recipe) for each growth section included in the entire cycle growth environment input value (full cycle cultivation recipe), the mentor cultivator 110 Monitor growth environment input values (cultivation recipe) for each growth section.

If an error occurs in the growth environment input value (cultivation recipe) for each step applied to the mass production cultivator 130, the mentor cultivator 110 can correct the growth environment input value (cultivation recipe) for each step.

The mentor cultivator 110 distributes different growth environment input values (cultivation recipes) to a plurality of mentee cultivators 120. Cultivation data, which is the result of cultivation by the mentee cultivators 120, is transmitted to the mentor cultivator 110 in real time.

The mentor cultivator 110 collects cultivation data from the mentee cultivator 120. The mentor cultivator 110 selects the growth environment input value (cultivation recipe) of a specific cultivator among the mentee cultivators 120 that are close to the cultivation purpose and then performs machine learning to group them. The mentor cultivator 110 selects the most dominant growth environment input (cultivation recipe) among each group and then directly cultivates them.

The mentor cultivator 110 performs a detailed analysis of the directly cultivated plants and determines the cultivation results by analyzing secondary metabolites (stress components of the plant).

The mentor cultivator 110 is a mentee cultivator when the cultivation result reaches the goal (secondary metabolites (stress components of plants) are more than a preset standard value or a recipe in which the secondary metabolites (stress components of plants) are generated the most). To (120), the growth environment input value (cultivation recipe) that is judged to have achieved the goal is transmitted. The mentee cultivator 120 determines that the goal received from the mentor cultivator 110 has been achieved and generates the next stage growth environment input value (cultivation recipe).

Figure 2 is a diagram schematically showing the internal modules of the mentee cultivator according to this embodiment.

The mentee cultivator 120 refers to a plant cultivator that is relatively cheaper than the mentor cultivator 110. The mentee cultivation unit 120 includes a plurality of cultivation units. Each of the plurality of cultivators included in the mentee cultivator 120 refers to a cultivator capable of cultivating various plants in large quantities using respective growth environment input values (cultivation recipes).

The mentee cultivator 120 cultivates various plants by performing light quantity control, cooling/heating control, CO2 control, FAN control, and nutrient solution supply control according to a plurality of growth environment input values (cultivation recipes).

The mentee cultivator 120 extracts the external pattern (outward appearance) of plants according to the growth process of the plants. The mentee cultivator 120 extracts an external pattern (outward appearance) based on the plant's RGB image, 3D scanning, and chlorophyll fluorescence measurement.

For example, the mentee cultivator 120 may include 150 cultivators. Of the 150 mentee cultivators (120), 75 can grow lettuce, and the remaining 75 can grow bok choy.

Figure 3 is a flowchart for explaining the AI-based plant cultivation method according to this embodiment.

The mentor cultivator 110 transmits the first growth environment input value (cultivation recipe) to the mentee cultivator 120 (S310).

The mentee cultivator 120 grows the plant according to the first growth environment input value, generates a first appearance pattern for the plant, and transmits it to the mentor cultivator 110 (S312). In step S312, the mentee cultivator 120 receives a plurality of different first growth environment input values (cultivation recipes) for each growth section divided by quantitative time units or the growth process in which plant leaves are generated from the mentor cultivator 110. Receive. The mentee cultivator 120 uses a plurality of cultivators to create a plurality of cultivation environments according to a plurality of different first growth environment input values (cultivation recipes), and grows plants in each of the plurality of cultivation environments. The mentee cultivator 120 uses a camera provided to generate a first external pattern (outward appearance) of a plant grown in a plurality of cultivation environments according to a plurality of first growth environment input values (cultivation recipes). The mentee cultivator 120 transmits the first cultivation data including the first appearance pattern (outward appearance) to the mentor cultivator 110.

The mentor cultivator 110 performs machine learning based on the first appearance pattern and creates a first group by first grouping by similar appearance (S314). In step S314, the mentor cultivator 110 performs machine learning based on the first appearance pattern (outward appearance) included in the first cultivation data received from the mentee cultivator 120 to first group by similar appearance to form a plurality of Create the first group of

The mentor cultivator 110 selects samples from the first group and extracts growth environment input values for each first sample for the samples (S316). In step S316, the mentor cultivator 110 selects one or more samples from each of the plurality of first groups. The mentor cultivator 110 extracts a growth environment input value (cultivation recipe) for each first sample.

The mentor cultivator 110 directly cultivates the plant using the growth environment input value for each first sample (S318). In step S318, the mentor cultivator 110 directly cultivates the plant for the first time using the growth environment input value (cultivation recipe) for each first sample.

The mentor cultivator 110 determines the first group ranking by detecting secondary metabolites of the directly cultivated plants (S320). In step S320, the mentor cultivator 110 uses the camera RGB camera, depth camera, and IR camera provided for the first directly cultivated plant to determine the total area of the leaf, the overall shape model, the chlorophyll of the leaf, and the temperature of the leaf as preset. By comparing with the reference value (area value in the absence of stress), secondary metabolites, which are stress components, are detected, and the most abundant secondary metabolites (plant stress) among the first plurality groups are detected for each growth period (planting, germination to harvest). Determine the rank of the first group in which the component) is detected.

The mentor cultivator 110 selects the second growth environment input value for the first upper group among the first group rankings and transmits it to the mentee cultivator 120 (S322). In step S322, the mentor cultivator 110 generates a large amount of secondary metabolites (stress components of plants) in the growth environment input value (cultivation recipe) corresponding to the first upper group within the preset ranking among the ranks of the first group. A plurality of second growth environment input values (cultivation recipes) are selected. The mentor cultivator 110 transmits a plurality of second growth environment input values (cultivation recipes) that enable the production of a large amount of secondary metabolites (plant stress components) back to the mentee cultivator 120.

The mentee cultivator 120 grows the plant according to the second growth environment input value, generates a second appearance pattern for the plant, and transmits it to the mentor cultivator 110 (S324). In step S324, the mentee cultivator 120 receives a plurality of second growth environment input values (cultivation recipes) that produce the largest amount of secondary metabolites (plant stress components) from the mentor cultivator 110. The mentee cultivator 120 uses a plurality of cultivators to create a plurality of cultivation environments according to a plurality of second growth environment input values (cultivation recipes) and grows plants in each of the cultivation environments. The mentee cultivator 120 uses an equipped camera to grow plants grown in a plurality of cultivation environments according to the second growth environment input value (cultivation recipe) that generates the most secondary metabolites (stress components of plants). 2 Create an external pattern (outward appearance). The mentee cultivator 120 transmits the Nth cultivation data including the second appearance pattern (outward appearance) to the mentor cultivator 110.

The mentor cultivator 110 performs machine learning based on the second appearance pattern and creates a second group by secondary grouping by similar appearance (S326). In step S326, the mentor cultivator 110 performs machine learning based on the second appearance pattern (outward appearance) included in the N-th cultivation data received from the mentee cultivator 120, and performs secondary grouping by similar appearance to form a plurality of Create a second group of

The mentor cultivator 110 selects samples from the second group and extracts growth environment input values for each second sample for the samples (S328). In step S328, the mentor cultivator 110 selects one or more samples from each of the plurality of second groups. The mentor cultivator 110 extracts a growth environment input value (cultivation recipe) for each second sample.

The mentor cultivator 110 directly cultivates the plant using the growth environment input value for each second sample (S330). In step S330, the mentor cultivator 110 directly cultivates the plant a second time using the growth environment input value (cultivation recipe) for each second sample.

The mentor cultivator 110 detects secondary metabolites of directly grown plants and determines the second group ranking (S332). In step S332, the mentor cultivator 110 measures the total area of the leaf, the overall shape model, the chlorophyll of the leaf, and the temperature of the leaf using the cameras provided for the plant, such as an RGB camera, a depth camera, and an IR camera. By comparing with a preset reference value (area value in the absence of stress), secondary metabolites, which are stress components, are detected and the most abundant secondary metabolites (plant The stress component of) determines the ranking of the detected second group.

The mentor cultivator 110 repeats the process of selecting the growth environment input value for the second higher group among the second group rankings (S334). In step S334, the mentor cultivator 110 generates a large amount of secondary metabolites (stress components of plants) in an environmental input value (cultivation recipe) corresponding to the second upper group within the preset ranking among the second group ranks. The process of selecting the optimal growth environment input value (cultivation recipe) is repeated to generate the optimal growth environment input value (cultivation recipe).

The mentor cultivator 110 selects a growth environment input value (cultivation recipe) that detects the largest amount of secondary metabolites (plant stress components) for the first section (seeding) of the growth sections. The mentor cultivator 110 selects the growth environment input value (cultivation recipe) that allows the detection of the most secondary metabolites (plant stress components) for the next section (germination) among the growth sections. The mentor cultivator 110 detects only the differential distribution of the growth environment input value (cultivation recipe) of the first section and the growth environment input value (cultivation recipe) of the next section.

The mentor cultivator 110 collects all optimal growth environment input values (cultivation recipes) for each growth section and generates full cycle growth environment input values (full cycle cultivation recipes) for the entire growth section. The mentor cultivator 110 transmits the full cycle growth environment input value (full cycle cultivation recipe) to the mass production cultivator 130.

The mentor cultivator 110 transmits the optimal growth environment input value selected through repetition to the mass production cultivator 130 (S336). In step S336, the mentor cultivator 110 transmits the optimal growth environment input value (cultivation recipe) to the mass production cultivator 130. The mass production cultivator 130 receives the optimal growth environment input value (cultivation recipe) from the mentor cultivator 110 using a plurality of cultivators and mass-produces plants. The mass production cultivator 130 extracts the growth environment input value (cultivation recipe) applied to each growth section included in the entire cycle growth environment input value (full cycle cultivation recipe). When the growth section changes, the mass production cultivator 130 mass-produces plants according to the growth environment input value (cultivation recipe) applied to the corresponding section.

When the mass production cultivator 130 applies the growth environment input value (cultivation recipe) for each growth section included in the entire cycle growth environment input value (full cycle cultivation recipe), the mentor cultivator 110 provides the growth environment for each growth section. If an error occurs while monitoring the input value (cultivation recipe), correct the growth environment input value (cultivation recipe) for each growth section.

In Figure 3, steps S310 to S336 are described as being sequentially executed, but this is not necessarily limited. In other words, the steps shown in FIG. 3 may be applied by modifying them or executing one or more steps in parallel, so FIG. 3 is not limited to a time-series order.

As described above, the AI-based plant cultivation method according to this embodiment shown in FIG. 3 may be implemented as a program and recorded on a computer-readable recording medium. A computer-readable recording medium on which a program for implementing the AI-based plant cultivation method according to this embodiment is recorded includes all types of recording devices that store data that can be read by a computer system.

Figure 4 is a diagram for explaining a growth section according to this embodiment.

The growth period can be divided into sowing period, germination period, cotyledon period, stem formation period, harvest period, etc. The interval between growth sections can be divided by preset time within 1 day.

The growth section can be divided into preset units of the entire growth time from sowing to harvest. The growth section can be divided into a growth process in which plant leaves are produced during the growth process, but is not necessarily limited to this, and can also be divided into quantitative time units (for example, 2 to 6 hours).

The growth and cultivation system according to this embodiment separates and manages the period from sowing and germination of plants to harvesting into preset time series units.

The mentor cultivator 110 distributes different first growth environment input values (cultivation recipes) to a plurality of mentee cultivators 120, respectively. The mentor cultivator 110 collects first cultivation data on plants grown by the mentee cultivators 120 based on each first growth environment input value (cultivation recipe) in real time (or at regular time intervals). The mentor cultivator 110 selects the first cultivation recipe that is closest to the cultivation purpose among the first cultivation data collected from the plurality of mentee cultivators 120. The mentor cultivator 110 directly cultivates plants in real time based on the first cultivation recipe that is closest to the cultivation purpose. The mentor cultivator 110 precisely analyzes cultivation data for directly grown plants and generates a second growth environment input value (cultivation recipe) that is closest to the cultivation purpose.

The mentor cultivator 110 distributes the second growth environment input value (cultivation recipe) closest to the cultivation purpose to a plurality of mentee cultivators 120. The mentor cultivator 110 collects second cultivation data on plants grown by the mentee cultivators 120 based on the second growth environment input value (cultivation recipe) in real time (or at regular time intervals). The mentor cultivator 110 selects a second cultivation recipe that is closest to the cultivation purpose among the second cultivation data collected from the plurality of mentee cultivators 120. The mentor cultivator 110 directly cultivates plants in real time based on the second cultivation recipe that is closest to the cultivation purpose. The mentor cultivator 110 precisely analyzes cultivation data for directly grown plants and generates a third growth environment input value (cultivation recipe) that is closest to the cultivation purpose.

As described above, the desired cultivation recipe is optimized within a short period of time by repeating the exchange of cultivation recipes between the mentor cultivator 110 and the mentee cultivator 120.

If the first cultivation data of the mentor cultivator 110 does not reach the target, the mentor cultivator 110 immediately cultivates with the cultivation recipe of another nearby mentee cultivator 120 to shorten the optimization cycle.

By using one Mentor cultivator (110), an expensive expert plant cultivator for precise analysis, multiple Menti cultivators (N units), which are low-priced plant cultivators, the data can be simultaneously increased by N times the cropping time of plant cultivation.

The mentor cultivator 110, the mentee cultivator 120, and the mass production cultivator 130 have the same protocol and operating system. When the optimal cultivation recipe is extracted from the relationship between the mentor cultivator 110 and the mentee cultivator 120, the mentor cultivator 110 transmits the optimal cultivation recipe to the mass production cultivator 130, and the corresponding plant (special purpose) is grown in the mass production cultivator 130. plants) can be mass-produced. When mass production cultivation begins in the mass production cultivation unit 130, the mentor cultivation unit 110 and the small number of mentee cultivation units 120 can also start cultivation of the same crop at the same time to synchronize cultivation.

The mentor cultivator 110 can monitor the growth data of the mass production cultivator 130 in real time, and compare and analyze the growth data collected by the mentee cultivator 120 in real time to manage whether the expected target value is met. If the mentor cultivator 110 deviates from the expected target value, it distributes the changed cultivation recipe to the mentee cultivator 120, and when the optimal value is obtained, it distributes the changed cultivation recipe to mass production to optimize plant cultivation for mass production in real time.

Figure 5 is a diagram for explaining the measurement of chlorophyll fluorescence of plants according to this embodiment.

The mentor cultivator 110 measures the fluorescence of chlorophyll in plants to detect stress in the cultivated plants. The mentor cultivator 110 transmits a plurality of light sources to the leaves of the plant via an aperture in order to measure the fluorescence of chlorophyll in the leaves of the plant. The mentor cultivator 110 measures chlorophyll fluorescence by receiving reflected light corresponding to the light source from the leaves of the plant via the aperture.

Chlorophyll fluorescence refers to a fluorescence phenomenon occurring in chlorophyll, a photosynthetic pigment in plants. Fluorescence generated from chlorophyll has the characteristic that chlorophyll absorbs short-wavelength blue or ultraviolet light and then emits long-wavelength red light. Although many pigments exhibit fluorescence, chlorophyll fluorescence can detect the photosynthesis process non-destructively and precisely.

The mentor cultivator 110 measures stomatal conductance and chlorophyll a fluorescence. Stomatal openings in plant leaves regulate the exchange of water vapor and CO2 between the leaf and the air, stomatal conductance to water (gsw), CO2 in response to light, temperature and humidity are measures of the degree of stomata opening and number of stomata. . Stomatal openings present in plant leaves can be used as an indicator of the plant's genetic makeup and physiological response to environmental conditions.

The mentor cultivator 110 provides information on the quantum efficiency, electron transport rate (ETR), and non-photochemical quenching (NPQ) of the leaf for measuring chlorophyll a fluorescence, as well as collectively protects the leaf when it absorbs excessive light energy. It senses information about various reactions.

Figure 6 is a diagram showing an RGB image according to this embodiment.

The mentor cultivator 110 senses the image of the plant. The Mentor cultivator 110 uses non-contact measurement to quickly and accurately sense plants for high-throughput phenotyping so that plants can find beneficial traits that enable them to resist abiotic and biotic stresses.

The mentor grower 110 senses images of plants and uses an RGB camera, a thermal imaging camera, and a 3D laser scanner. The mentor cultivator 110 acquires an RGB color image using an RGB camera.

The mentor cultivator 110 uses a thermal imaging camera to measure the temperature of leaves and plants using the temperature of an object photographed with infrared rays. The Mentor grower 110 uses a 3D laser scanner to accurately and structurally represent plants and create 3D models. The mentor cultivator 110 measures chlorophyll fluorescence using a fluorescence image captured by a camera. The Mentor grower 110 generates 3D hyperspectral data of plants in pixel units in the spectral range of 400 to 2500 nm based on the spectral image.

The mentor cultivator 110 detects special components (plant stress components) for cultivated plants using RGB color images, thermal images, 3D images, fluorescence images, and spectral images.

The mentor cultivator 110 measures the physiological state of the plant using RGB color images, thermal images, 3D images, fluorescence images, and spectral images. The mentor cultivator 110 measures the growth and development status of plants using RGB color images, thermal images, 3D images, fluorescence images, and spectral images. The Mentor grower 110 measures yield and biomass production using RGB color images, thermal images, 3D images, fluorescence images, and spectral images. The mentor cultivator 110 measures tolerance and resistance to biotic and abiotic stresses using RGB color images, thermal images, 3D images, fluorescence images, and spectral images. The mentor cultivator 110 expresses the structure of the plant using RGB color images, thermal images, 3D images, fluorescence images, and spectral images.

The mentor cultivator 110 detects plant growth and development over time using digital color RGB imaging or 3D scanning technology. The mentor grower 110 measures high-resolution movement of visible characteristics to extract plant shape, structure, and color index characteristics. The mentor cultivator 110 uses RGB color images to calculate the area, height, and growth width of the plant. The mentor cultivator 110 uses RGB color images to extract the number of leaves, evaluate biomass, calculate relative growth rate, and generate leaf movement, combination plan view, and side view.

Figure 7 is a diagram showing 3D scanning according to this embodiment.

The mentor grower 110 uses a 3D laser scanner to express structural plants and create a 3D model. The mentor grower 110 performs 3D scanning on plants using a 3D laser scanner. The mentor grower 110 uses a 3D laser scanner to express the structure of the plant and provide 3D modeling.

Figure 8 is a diagram showing an example of chlorophyll fluorescence measurement according to this embodiment.

The mentor cultivator 110 measures the chlorophyll fluorescence of the plant's leaves and then generates information about the plant's photosynthetic ability, physiological and metabolic state, and sensitivity to various stress conditions.

The above description is merely an illustrative explanation of the technical idea of the present embodiment, and those skilled in the art will be able to make various modifications and variations without departing from the essential characteristics of the present embodiment. Accordingly, the present embodiments are not intended to limit the technical idea of the present embodiment, but rather to explain it, and the scope of the technical idea of the present embodiment is not limited by these examples. The scope of protection of this embodiment should be interpreted in accordance with the claims below, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of rights of this embodiment.

110: Mentor Cultivation Machine
120: Mentee cultivation period
130: Parasol cultivation machine

Claims (16)

A mentee cultivator that cultivates plants according to a plurality of different first growth environment input values using a plurality of cultivators and generates first cultivation data for the plants;
Among the first cultivation data, the growth environment input value closest to the cultivation purpose is selected, the plant is directly cultivated based on the growth environment input value closest to the cultivation purpose, and the cultivation data for the directly cultivated plant is analyzed. Generate a plurality of second growth environment input values based on the results,
After transmitting the plurality of second growth environment input values back to the mentee cultivator, N-th cultivation data corresponding to the second growth environment input value is received from the mentee cultivator, and among the N-th cultivation data, the most suitable for cultivation purposes is received. Select the closest growth environment input value, directly cultivate the plant based on the growth environment input value closest to the cultivation purpose, and repeat the process of analyzing the cultivation data for the directly cultivated plant to determine the optimal growth environment input value. A mentor cultivator that creates;
A mass production cultivator that mass-produces the plants by receiving the optimal growth environment input value from the mentor cultivator using a plurality of cultivators;
A plant cultivation system comprising: According to paragraph 1,
A plant cultivation system, characterized in that the mentor cultivator, the mentee cultivator, and the mass production cultivator communicate with each other using the same protocol and the same operating system. According to paragraph 1,
The mentee cultivation period is
A plurality of different first growth environment input values are received from the mentor cultivator for each growth process in which leaves of the plant are generated or growth sections divided by quantitative time units, and a plurality of different first growth environment input values are received using a plurality of cultivators. A plant cultivation system comprising creating a plurality of cultivation environments according to a first growth environment input value and growing the plants in each of the plurality of cultivation environments. According to paragraph 3,
The mentee cultivation period is
Using a provided camera, a first external pattern of the plant grown in the plurality of cultivation environments is generated according to the plurality of first growth environment input values, and the first cultivation data including the first external pattern are generated. A plant cultivation system characterized in that transmission to the mentor cultivator. According to paragraph 4,
The mentor cultivation period
Machine learning is performed based on the first appearance pattern included in the first cultivation data received from the mentee cultivator to create a plurality of first groups by first grouping by similar appearance, and the plurality of Selecting one or more samples from each of the first groups, extracting growth environment input values for each first sample for the samples, and directly cultivating the plants using the growth environment input values for each first sample. Plant cultivation system. According to clause 5,
The mentor cultivation period
The total area of the leaf, the overall shape model, the chlorophyll of the leaf, and the temperature of the leaf are compared with preset reference values using cameras such as an RGB camera, a depth camera, and an IR camera provided for the plants grown directly. A plant cultivation system characterized by detecting metabolites and determining the ranking of the first group in which the most secondary metabolites were detected among the plurality of first groups for each growth section. According to clause 6,
The mentor cultivation period
The growth environment input value corresponding to the first upper group within the preset ranking among the ranks of the first group is selected as the plurality of second growth environment input values that cause the secondary metabolites to be generated in large quantities, and the secondary metabolism A plant cultivation system characterized in that a plurality of input values of the second growth environment for producing a large number of products are transmitted back to the mentee cultivator. In clause 7,
The mentee cultivation period is
Receiving a plurality of second growth environment input values that produce the most secondary metabolites from the mentor cultivator, and creating a plurality of cultivation environments according to the plurality of second growth environment input values using a plurality of cultivators. A plant cultivation system characterized by creating and growing the plants in each of the cultivation environments. According to clause 8,
The mentee cultivation period is
Using a provided camera, a second appearance pattern of the plants grown in a plurality of cultivation environments is generated according to the second growth environment input value that produces the most secondary metabolites, and the second appearance pattern A plant cultivation system, characterized in that transmitting the Nth cultivation data including to the mentor cultivator. According to clause 9,
The mentor cultivation period
Machine learning is performed based on the second appearance pattern included in the N-th cultivation data received from the mentee cultivator to create a plurality of second groups by secondary grouping by similar appearance, and each of the plurality of second groups Plant cultivation, characterized in that one or more samples are selected from the sample, a growth environment input value for each second sample is extracted from the sample, and the plant is directly cultivated in a secondary manner using the growth environment input value for each second sample. system. According to clause 10,
The mentor cultivation period
For the plants grown directly, the total area of the leaves, the overall shape model, the chlorophyll of the leaves, and the temperature of the leaves are compared with preset reference values using the camera RGB camera, depth camera, and IR camera provided for the plant to determine stress. A plant cultivation system characterized by detecting secondary metabolites, which are components, and determining the ranking of the second group in which the most secondary metabolites were detected among the plurality of second groups for each growth section. According to clause 11,
The mentor cultivation period
Inputting the optimal growth environment by repeating the process of selecting the environmental input value corresponding to the second upper group within the preset ranking among the rankings of the second group as the optimal growth environment input value for producing a large amount of secondary metabolites. A plant cultivation system characterized in that it generates a value and transmits the optimal growth environment input value to the mass production cultivator. According to clause 6,
The mentor cultivation period
After selecting a growth environment input value that causes the most secondary metabolites to be detected for the first section of the growth sections, select a growth environment input value that causes the most secondary metabolites to be detected for the next section of the growth sections. A plant cultivation system characterized in that, after selection, only the differential distribution of the growth environment input value of the first section and the growth environment input value of the next section is detected. According to clause 13,
The mentor cultivation period
A plant cultivation system characterized by collecting all the optimal growth environment input values for each growth section, generating a full cycle growth environment input value for the entire growth section, and transmitting the entire cycle growth environment input value to the mass production cultivator. . According to clause 14,
The mass production cultivation machine
Characterized in that the growth environment input value applied to each growth section included in the overall cycle growth environment input value is extracted, and when the growth section is changed, the plants are mass-produced according to the growth environment input value applied to the section. Plant cultivation system. According to clause 14,
The mentor cultivation period
When the mass production grower applies the growth environment input value for each growth section included in the overall cycle growth environment input value, if an error occurs while monitoring the growth environment input value for each growth section, the growth environment input value for each growth section is input. A plant cultivation system characterized by allowing the value to be modified.