A sprout growth monitoring method and device based on multi-parameter sensing

Abstract

This application discloses a method and device for monitoring bean sprout growth based on multi-parameter sensing, comprising: acquiring environmental parameters and growth images during the bean sprout growth process through multiple sensors and image acquisition devices; preprocessing the environmental parameters and growth images to extract environmental and image recognition features; inputting the environmental and image recognition features into a pre-trained growth stage prediction model to predict the growth stage and nutritional indicators of the bean sprouts; determining whether the bean sprouts are in the optimal harvesting period based on the predicted growth stage and nutritional indicators, and generating harvesting suggestions; when the predicted growth stage is the nutrient enrichment period and the nutritional indicators reach a preset threshold, determining that the optimal harvesting period has been reached, and pushing a harvesting reminder to the user. This application can accurately predict the peak nutritional value and optimal harvesting period of bean sprouts, and only requires low-cost commercial sensors, without the need for expensive equipment such as GC-MS, making it truly suitable for home use.

Description

Technical Field

[0001] This application relates to the field of bean sprout growth monitoring technology, specifically to a method and device for monitoring bean sprout growth based on multi-parameter sensing. Background Technology

[0002] With "healthy eating" and "sustainable development" becoming the dominant themes of our time, food safety and nutrition management are gradually becoming core issues of social concern. Bean sprouts, a common vegetable in Chinese and East Asian households, are considered an ideal health food due to their high protein, low fat, and rich vitamin and dietary fiber content.

[0003] With social development and technological advancements, modern consumers' dietary concepts are undergoing significant changes. A healthy, natural, and sustainable lifestyle is gradually replacing the simple demands of "eating enough" and "convenience," with people desiring a greater sense of control and participation in the eating process. The home kitchen has thus become a new scenario for technology-enabled healthy living. More and more consumers are choosing to make their own bean sprouts at home, which is both a proactive response to food safety concerns and reflects a lifestyle of "returning to nature and embracing science."

[0004] Against this backdrop, how to leverage technology to achieve safe, controllable, nutritionally optimized, and intelligent management of family food ingredients has become an important issue for the future of healthy eating.

[0005] A patent document with publication number CN118171219A discloses a method for monitoring the growth status of mung bean sprouts, applied to a nutrient substrate incubator. The method includes: acquiring growth status characteristic parameters of mung bean sprouts, wherein the growth status characteristic parameters include color information, gas parameter information, and body surface information; acquiring color information and gloss information of mung bean sprouts based on color information, acquiring the appearance color of mung bean sprouts based on color information and gloss information, and extracting numerical values ​​of the appearance color of mung bean sprouts based on the LAB color space to obtain the appearance color growth coefficient. The VOC and ammonia content of the air in the growth area of ​​mung bean sprouts are obtained based on gas parameters. This is achieved using gas chromatography-mass spectrometry (GC-MS), and abnormal air quality molecular information is acquired based on the VOC and ammonia content. The content coefficients of these molecules are used as gas molecular coefficients. The surface temperature and humidity of the mung bean sprouts are obtained based on their surface information, and a comprehensive body temperature growth coefficient is derived. The appearance color growth coefficient, gas growth coefficient, and comprehensive body temperature growth coefficient are input into a mung bean sprout growth state model for training, resulting in a growth state coefficient. It is then determined whether this growth state coefficient falls within a preset range. If the growth state coefficient is outside this range, a corresponding correction command is generated. The growth state of the mung bean sprouts is then controlled based on the correction command.

[0006] However, this method has the following drawbacks:

[0007] 1. The “growth morphology coefficient” in this method is essentially a simple weighted sum (or product) of appearance color growth coefficient, gas growth coefficient, and comprehensive body temperature growth coefficient. It belongs to a linear combination model and lacks the ability to model nonlinear relationships, which makes it impossible to accurately determine the optimal harvesting period and predict the nutritional components of bean sprouts (vitamin C, polyphenols, flavonoids, etc.).

[0008] 2. This method relies on gas chromatography-mass spectrometry (GC-MS) to detect VOCs and ammonia, but the equipment is expensive (hundreds of thousands to millions of yuan), complex to operate, and requires professional maintenance, making it completely unsuitable for home use. Summary of the Invention

[0009] Therefore, this application provides a method and device for monitoring bean sprout growth based on multi-parameter sensing to solve the problems that existing technologies cannot accurately predict the optimal harvest period and nutritional components, and are difficult to apply to home settings.

[0010] To achieve the above objectives, this application provides the following technical solution:

[0011] Firstly, a method for monitoring bean sprout growth based on multi-parameter sensing includes:

[0012] Environmental parameters and growth images during the bean sprout growth process are acquired through multiple sensors and image acquisition devices; the multiple sensors include: temperature and humidity sensor, pH sensor, water quality sensor, soil moisture sensor and photosensitizer;

[0013] The environmental parameters and the growth image are preprocessed, and environmental features and image recognition features are extracted. The environmental features include at least one of the following: temperature-humidity product, average temperature, average humidity, standard deviation of temperature, standard deviation of humidity, average light intensity, and average total dissolved solids.

[0014] The environmental features and the image recognition features are input into a pre-trained growth stage prediction model to predict the current growth stage and nutritional indicators of the bean sprouts. The growth stages include: initial stage, rapid growth stage, nutrient accumulation stage, and harvest stage. The nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols.

[0015] Based on the predicted growth stage and nutritional indicators, determine whether it is the optimal harvesting period and generate harvesting recommendations.

[0016] When the predicted growth stage is the nutrient enrichment period and the nutrient indicators reach the preset threshold, the optimal harvest period is determined, and a harvest reminder is pushed to the user.

[0017] Preferably, the data preprocessing for the environmental features includes: removing outliers from the environmental parameters; and using Kalman filtering to filter the removed data to obtain stable, continuous data.

[0018] Preferably, the growth stage prediction model is trained using a random forest model.

[0019] Preferably, the growth stage prediction model predicts growth stage and nutritional indicators based on the mapping relationship between environment and nutrition learned by compressed dictionary.

[0020] Preferably, the mapping relationship between environment and nutrition is as follows: when the ambient temperature is maintained at 25-30℃, the humidity is maintained at 70%-90%, and the pH value is maintained at 6.0-7.0, the bean sprouts' growth capacity reaches its optimal state, and the nutrient content reaches the optimal control value of over 90%.

[0021] Preferably, the bean sprouts are 4-5 cm long when harvested at the optimal time, when their polyphenol and antioxidant activities reach their peak.

[0022] Preferably, an abnormal alarm procedure is also included: when the humidity is >95% and the temperature is >28°C for more than 2 hours, a high mold risk alarm is triggered; when the pH value is consistently below 5.5, a root metabolism abnormality alarm is triggered.

[0023] Secondly, a bean sprout growth monitoring device based on multi-parameter sensing includes:

[0024] The data acquisition module is used to acquire environmental parameters and growth images during the growth process of bean sprouts through multiple sensors and image acquisition devices; the multiple sensors include: temperature and humidity sensor, pH sensor, water quality sensor, soil moisture sensor and photosensor;

[0025] The data preprocessing module is used to preprocess the environmental parameters and the growth image, and extract environmental features and image recognition features; the environmental features include at least one of the following: temperature-humidity product, average temperature, average humidity, standard deviation of temperature, standard deviation of humidity, average light intensity, and average total dissolved solids.

[0026] The growth stage prediction module is used to input the environmental features and the image recognition features into a pre-trained growth stage prediction model to predict the current growth stage and nutritional indicators of the bean sprouts; the growth stages include: initial stage, rapid growth stage, nutrient enrichment stage and harvest stage; the nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols.

[0027] The harvest suggestion generation module is used to determine whether the optimal harvesting period is reached based on the predicted growth stage and nutritional indicators, and to generate harvest suggestions.

[0028] The harvest reminder module is used to determine the optimal harvest time when the predicted growth stage is the nutrient enrichment period and the nutrient indicators reach the preset threshold, and to push a harvest reminder to the user.

[0029] Thirdly, a computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of a method for monitoring the growth of bean sprouts based on multi-parameter sensing.

[0030] Fourthly, a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for monitoring the growth of bean sprouts based on multi-parameter sensing.

[0031] Compared with the prior art, this application has at least the following beneficial effects:

[0032] This application provides a method for monitoring bean sprout growth based on multi-parameter sensing, comprising: acquiring environmental parameters and growth images during the bean sprout growth process through multiple sensors and image acquisition devices; preprocessing the environmental parameters and growth images to extract environmental features and image recognition features; inputting the environmental features and image recognition features into a pre-trained growth stage prediction model to predict the current growth stage and nutritional indicators of the bean sprouts; the growth stages include: initial stage, rapid growth stage, nutrient enrichment stage, and harvesting stage; nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols; determining whether the optimal harvesting period is reached based on the predicted growth stage and nutritional indicators, and generating harvesting suggestions; when the predicted growth stage is the nutrient enrichment stage and the nutritional indicators reach a preset threshold, determining that the optimal harvesting period has been reached, and pushing a harvesting reminder to the user. This application can accurately predict the peak nutritional value and optimal harvesting period of bean sprouts, and only requires low-cost commercial sensors, without the need for expensive equipment such as GC-MS, making it truly suitable for home scenarios. Attached Figure Description

[0033] To more intuitively illustrate the prior art and this application, exemplary drawings are provided below. It should be understood that the specific shapes and structures shown in the drawings should not generally be regarded as limiting conditions for implementing this application; for example, based on the technical concept disclosed in this application and the exemplary drawings, those skilled in the art are able to easily make conventional adjustments or further optimizations to the addition / reduction / classification, specific shapes, positional relationships, connection methods, size ratios, etc. of certain units (components).

[0034] Figure 1A flowchart of a method for monitoring bean sprout growth based on multi-parameter sensing provided in Embodiment 1 of this application;

[0035] Figure 2 This is a schematic diagram of a bean sprout growth monitoring system based on multi-parameter sensing, provided in Embodiment 5 of this application. Detailed Implementation

[0036] The present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] In the description of this application: unless otherwise stated, "a plurality of" means two or more. The terms "first," "second," "third," etc., in this application are intended to distinguish the objects referred to and do not have any special meaning in terms of technical connotation (e.g., they should not be construed as an emphasis on importance or order). Expressions such as "including," "comprising," and "having" also mean "not limited to" (certain units, components, materials, steps, etc.).

[0038] The terms used in this application, such as "upper," "lower," "left," "right," and "middle," are generally used to indicate the general relative positional relationship for the purpose of intuitive understanding by referring to the accompanying drawings, and are not absolute limitations on the positional relationship in the actual product.

[0039] Example 1

[0040] Please see Figure 1 This embodiment provides a method for monitoring bean sprout growth based on multi-parameter sensing, including:

[0041] S1: Acquire environmental parameters and growth images during the bean sprout growth process through multiple sensors and image acquisition devices; the multiple sensors include: temperature and humidity sensor, pH sensor, water quality sensor, soil moisture sensor, and photosensor.

[0042] Specifically, this embodiment uses an SHT31 temperature and humidity sensor, an ADS1115 water quality probe (i.e., a water quality sensor), and a pH sensor to collect environmental parameters in real time, including temperature, humidity, pH value, and conductivity. Data is collected every 5 minutes, stored in a local buffer, and then uploaded to the cloud via WiFi+MQTT.

[0043] S2: Perform data preprocessing on environmental parameters and growth images, and extract environmental features and image recognition features; environmental features include at least one of the following: temperature-humidity product, average temperature, average humidity, standard deviation of temperature, standard deviation of humidity, average light intensity, and average total dissolved solids (i.e., average TDS);

[0044] Specifically, the preprocessing of environmental parameters includes: correcting and equalizing the collected temperature, humidity, water quality, and pH data, and removing constant values ​​(such as temperatures exceeding the 0-50°C range or pH values ​​exceeding the 0-14 range). Then, noise reduction is performed using wavelet filtering and Kalman filtering algorithms to eliminate bursty data caused by physical measurement market price and network interference, thereby removing interference signals caused by network interference and physical measurement interference to the greatest extent possible, and obtaining more stable and continuously updated temperature, humidity, and pH data.

[0045] Based on the preprocessed environmental parameters, the temperature-humidity product, average temperature, average humidity, standard deviation of temperature, and standard deviation of humidity are calculated as relevant feature vectors to more comprehensively describe the environmental state. Among them, the temperature-humidity product (T×H) represents the degree of complexity of the operating environment, which has a significant impact on the germination rate of mung beans; the average temperature and humidity reflect the basic state of the environment; and the standard deviation of temperature and humidity represents the stability of the environment, recording the fluctuations of the environment during the growth of mung beans.

[0046] When performing data preprocessing on growth images and extracting image recognition features, the process includes grayscale conversion, noise reduction, contour extraction, and extraction of image recognition features such as bud length, thickness, plant height, uniformity, color, and morphological density.

[0047] S3: Input environmental features and image recognition features into a pre-trained growth stage prediction model to predict the current growth stage and nutritional indicators of bean sprouts; growth stages include: initial stage, rapid growth stage, nutrient accumulation stage and harvest stage; nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols.

[0048] Specifically, the growth stage prediction model was trained using a random forest model (which effectively handles nonlinear relationships and has strong anti-overfitting ability). To obtain the core dataset required for model training, this embodiment conducted multiple sets of experiments to explore the influence of environmental variables on the bean sprout germination process, including the effects of key factors such as different temperatures (20, 25, 30°C), humidity (70%, 80%, 90%), and pH values ​​(6.0, 6.5, 7.0) on the germination rate and growth rate of mung bean sprouts.

[0049] To transform experimental data and theory into practically applicable prediction and control algorithms, this embodiment cleans, denoises, and extracts features (such as temperature and humidity fluctuation ranges) from the collected time-series data. Then, a random forest algorithm is used, with environmental data as input and growth status and nutrient tags as output, to train the prediction model. Finally, the model is evaluated and optimized to achieve intelligent decision-making.

[0050] In other words, the growth stage prediction model in this embodiment is trained based on an "environment-nutrient" dataset obtained through literature and experimental research methods. This dataset contains actual nutrient measurements of bean sprouts at various growth stages under different environmental conditions. By inputting the real-time collected environmental data into the trained random forest model, the model can output a prediction of the nutritional quality of bean sprouts in the next 12-24 hours, thereby determining whether the current stage is within the optimal harvest window.

[0051] In this embodiment, the growth stage prediction model predicts the growth stage and nutritional indicators based on the mapping relationship between environment and nutrition learned from the compressed dictionary. Specifically, the mapping relationship between environment and nutrition is as follows: when the ambient temperature is maintained at 25-30℃, the humidity is maintained at 70%~90%, and the pH value is maintained at 6.0-7.0, the bean sprouts reach their optimal growth capacity, and the nutritional components (such as vitamin C, polyphenols, etc.) reach more than 90% of the optimal control value.

[0052] In this embodiment, the growth stage prediction model uses environmental features and image recognition features as dual inputs. By combining environmental conditions with the actual appearance of the bean sprouts, it can achieve more accurate prediction of growth stage and nutritional indicators.

[0053] S4: Based on the predicted growth stage and nutritional indicators, determine whether it is the optimal harvest period and generate harvesting recommendations;

[0054] Specifically, at the optimal harvest time, bean sprouts are 4-5cm long, and their polyphenol and antioxidant activities reach their peak.

[0055] S5: When the predicted growth stage is the nutrient enrichment period and the nutrient indicators reach the preset threshold, the optimal harvest period is determined, and a harvest reminder is pushed to the user.

[0056] Specifically, when the model predicts that the mung beans have reached the "nutrient enrichment period" and the indicators meet the harvesting conditions, a "best harvesting time has been reached" push notification is sent through the cloud-app to remind users to harvest immediately.

[0057] This embodiment provides a method for monitoring bean sprout growth based on multi-parameter sensing, which also includes an abnormal alarm step: when the humidity is >95% and the temperature is >28°C for more than 2 hours, a high mold risk alarm is triggered; when the pH value is continuously below 5.5, a root metabolism abnormality alarm is triggered.

[0058] Specifically, an abnormal alarm is triggered by a germination abnormality warning model, which can identify risks such as rot and mold in advance. The germination abnormality warning model is a rule-based expert system. The rules include: (1) when the humidity is >95% and the temperature is >28°C for 2 consecutive hours, a "high mold risk" alarm is triggered; (2) when the pH value continues to drop and exceeds the normal range (e.g., <5.5), a "root metabolism abnormality" alarm is triggered; (3) it runs in parallel with the prediction model to provide users with timely risk intervention guidance.

[0059] The following experimental data further validates the bean sprout growth monitoring method based on multi-parameter sensing provided in this embodiment.

[0060] 1. Experimental groups: (1) Experimental group: Bean sprouts germinating using the method provided in this embodiment for monitoring and guidance. (2) Control group: Bean sprouts germinating according to conventional home experience methods.

[0061] 2. Experimental conditions: Both groups used the same batch of mung bean seeds, the same water source and germination container, and conducted a germination experiment for 21 days in the same indoor environment, with each group repeated 3 times.

[0062] The key indicators for the experimental group and the control group are shown in Table 1:

[0063] Table 1

[0064]

[0065] Results analysis:

[0066] 1. Reliability: The germination success rate of the experimental group was high and stable, and the decay rate was extremely low, which proved the effectiveness of the method in maintaining the optimal germination environment and preventing diseases.

[0067] 2. Bean sprout quality: The length and thickness of the radicle of the bean sprouts in the experimental group were significantly higher than those in the control group. This directly verifies that the harvesting recommendations based on algorithm prediction can accurately capture the growth peak. At the same time, based on the optimal peak period of nutrition in literature research, it helps users maximize the nutritional value.

[0068] 3. User experience: The complex germination process is simplified into clear data and explicit instructions, which greatly reduces the user's operational difficulty and psychological anxiety, and achieves the design goal of "peace of mind and ease".

[0069] The bean sprout growth monitoring method based on multi-parameter sensing provided in this embodiment has the following advantages:

[0070] 1. Accurate prediction of nutrient peak and optimal harvest period: This embodiment uses nonlinear machine learning algorithms such as random forest to effectively fit the complex nonlinear relationship between environmental parameters (temperature, humidity, pH, etc.) and the nutrient components of bean sprouts (vitamin C, polyphenols, etc.), accurately capture the peak change pattern of nutrient components that first rise and then fall, thereby achieving intelligent and accurate prediction of the optimal harvest period, solving the problem that existing linear models cannot predict nutrient dynamics.

[0071] 2. Low cost and suitable for home use: This embodiment uses only low-cost commercial components such as the SHT31 temperature and humidity sensor and the ADS1115 pH sensor, without relying on expensive equipment such as gas chromatography-mass spectrometry (GC-MS). The overall hardware cost is controlled at the hundred-yuan level. It is simple to operate and easy to maintain, and truly realizes intelligent monitoring of bean sprouts in home settings.

[0072] 3. High model accuracy and robustness: The random forest algorithm has the advantages of strong anti-overfitting ability and the ability to handle high-dimensional features (such as temperature and humidity product, fluctuation variance, etc.). Compared with the existing linear weighted summation model, the prediction accuracy is significantly improved.

[0073] Example 2

[0074] This embodiment provides a bean sprout growth monitoring device based on multi-parameter sensing, including:

[0075] The data acquisition module is used to acquire environmental parameters and growth images during the growth process of bean sprouts through multiple sensors and image acquisition devices; the multiple sensors include: temperature and humidity sensor, pH sensor, water quality sensor, soil moisture sensor and photosensor;

[0076] The data preprocessing module is used to preprocess the environmental parameters and growth images, and extract environmental features and image recognition features; the environmental features include at least one of the following: temperature-humidity product, average temperature, average humidity, standard deviation of temperature, standard deviation of humidity, average light intensity, and average total dissolved solids.

[0077] The growth stage prediction module is used to input the environmental features and image recognition features into a pre-trained growth stage prediction model to comprehensively predict the current growth stage and nutritional indicators of the bean sprouts; the growth stages include: initial stage, rapid growth stage, nutrient enrichment stage and harvest stage; the nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols.

[0078] The harvest suggestion generation module is used to determine whether the optimal harvesting period is reached based on the predicted growth stage and nutritional indicators, and to generate harvest suggestions.

[0079] The harvest reminder module is used to determine the optimal harvest time when the predicted growth stage is the nutrient enrichment period and the nutrient indicators reach the preset threshold, and to push a harvest reminder to the user.

[0080] For details on the implementation of each module in a bean sprout growth monitoring device based on multi-parameter sensing, please refer to the above description of the limitations of a bean sprout growth monitoring method based on multi-parameter sensing, which will not be repeated here.

[0081] Example 3

[0082] This embodiment provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of a method for monitoring the growth of bean sprouts based on multi-parameter sensing.

[0083] Example 4

[0084] This embodiment provides a computer-readable storage medium storing a computer program, which, when executed by a processor, represents the steps of a method for monitoring the growth of bean sprouts based on multi-parameter sensing.

[0085] Example 5

[0086] Please see Figure 2 This embodiment provides a bean sprout growth monitoring system based on multi-parameter sensing. The system acts as a "smart brain," responsible for processing data collected by the hardware, executing intelligent algorithms, and interacting with the user. The system adopts a layered architecture design, including embedded firmware, cloud services, and mobile applications.

[0087] I. Embedded Firmware

[0088] Development environment: Based on STM32, developed using C / C++ language.

[0089] Main tasks: (1) Sensor driving: Write the underlying drivers for devices such as temperature and humidity sensor (SHT31) and pH sensor (ADS1115) to achieve accurate data acquisition; (2) Task scheduling: Use FreeRTOS real-time operating system to create multiple independent tasks (such as data acquisition task, communication task, status monitoring task) to ensure timely system response and stable operation; (3) Data preprocessing and uploading: Filter (such as Kalman filtering) and calibrate the collected raw data, and stably upload the data packets to the mobile phone through the Wi-Fi module.

[0090] II. Cloud Services

[0091] Architecture: The cloud backend is built using AWS IoT Core or Alibaba Cloud IoT Platform as the core.

[0092] Core functions: (1) Data access and storage: The IoT platform is responsible for receiving and verifying the data uploaded by the embedded terminal, and then storing it in a time series database (such as InfluxDB) for easy time series analysis; (2) Intelligent model deployment: A trained random forest prediction model is deployed in the cloud. When new environmental data is received, the model will automatically trigger inference to predict the current growth stage index of the bean sprouts; (3) Business logic and alarms: Based on the prediction results and the set rules (such as continuous high humidity and high temperature), the cloud service will generate corresponding "best harvesting suggestions" or "abnormal status alarms" and send them to users in real time through the cloud-App push channel.

[0093] III. Mobile Application Design

[0094] Development framework: React Native or WeChat Mini Program framework is used to achieve cross-platform development.

[0095] Core Interface and Functions: (1) Data Dashboard: The main interface displays the temperature, humidity, pH value change trends and nutrient prediction results in real time in the form of curves, dashboards and other visualizations; (2) Intelligent Alert Center: It centrally displays all harvesting suggestions and abnormal alarms issued by the system, and users can view the handling solutions with one click; (3) Knowledge Base and History: It has a built-in knowledge base for bean sprout germination and supports viewing historical germination records to help users accumulate experience.

[0096] The technical features of the above embodiments can be combined in any way (as long as there is no contradiction in the combination of these technical features). For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described; these embodiments not explicitly written should also be considered to be within the scope of this specification.

Claims

1. A method for monitoring bean sprout growth based on multi-parameter sensing, characterized in that, include: Environmental parameters and growth images of bean sprouts during their growth process were acquired using multiple sensors and image acquisition devices. The plurality of sensors include: a temperature and humidity sensor, a pH sensor, a water quality sensor, a soil moisture sensor, and a photosensitizer; The environmental parameters and the growth image are preprocessed, and environmental features and image recognition features are extracted. The environmental features include at least one of the following: temperature-humidity product, average temperature, average humidity, standard deviation of temperature, standard deviation of humidity, average light intensity, and average total dissolved solids. The environmental features and the image recognition features are input into a pre-trained growth stage prediction model to comprehensively predict the current growth stage and nutritional indicators of the bean sprouts; the growth stages include: initial stage, rapid growth stage, nutrient accumulation stage and harvest stage; the nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols. Based on the predicted growth stage and nutritional indicators, determine whether it is the optimal harvesting period and generate harvesting recommendations. When the predicted growth stage is the nutrient enrichment period and the nutrient indicators reach the preset threshold, the optimal harvest period is determined, and a harvest reminder is pushed to the user.

2. The method for monitoring bean sprout growth based on multi-parameter sensing according to claim 1, characterized in that, The data preprocessing for the environmental features includes: removing outliers from the environmental parameters; and using Kalman filtering to filter the removed data to obtain stable, continuous data.

3. The method for monitoring bean sprout growth based on multi-parameter sensing according to claim 1, characterized in that, The growth stage prediction model was trained using a random forest model.

4. The method for monitoring bean sprout growth based on multi-parameter sensing according to claim 1, characterized in that, The growth stage prediction model predicts growth stages and nutritional indicators based on the mapping relationship between environment and nutrition learned from the compressed dictionary.

5. The method for monitoring bean sprout growth based on multi-parameter sensing according to claim 1, characterized in that, The mapping relationship between environment and nutrition is as follows: when the ambient temperature is maintained at 25-30℃, the humidity is maintained at 70%-90%, and the pH value is maintained at 6.0-7.0, the growth capacity of bean sprouts reaches the optimal state, and the nutrient content reaches more than 90% of the optimal control value.

6. The method for monitoring bean sprout growth based on multi-parameter sensing according to claim 1, characterized in that, At the optimal harvest time, the bean sprouts are 4-5 cm long, and their polyphenol and antioxidant activities reach their peak.

7. The method for monitoring bean sprout growth based on multi-parameter sensing according to claim 1, characterized in that, It also includes abnormal alarm steps: when the humidity is >95% and the temperature is >28°C for more than 2 hours, a high mold risk alarm is triggered; when the pH value is consistently below 5.5, a root metabolism abnormality alarm is triggered.

8. A bean sprout growth monitoring device based on multi-parameter sensing, characterized in that, include: The data acquisition module is used to acquire environmental parameters and growth images of bean sprouts during their growth process through multiple sensors and image acquisition devices; The plurality of sensors include: a temperature and humidity sensor, a pH sensor, a water quality sensor, a soil moisture sensor, and a photosensitizer; The data preprocessing module is used to preprocess the environmental parameters and the growth image, and extract environmental features and image recognition features; the environmental features include at least one of the following: temperature-humidity product, average temperature, average humidity, standard deviation of temperature, standard deviation of humidity, average light intensity, and average total dissolved solids. The growth stage prediction module is used to input the environmental features and the image recognition features into a pre-trained growth stage prediction model to comprehensively predict the current growth stage and nutritional indicators of the bean sprouts; the growth stages include: initial stage, rapid growth stage, nutrient accumulation stage and harvest stage; the nutritional indicators include the relative content of vitamin C and / or the enrichment level of polyphenols. The harvest suggestion generation module is used to determine whether the optimal harvesting period is reached based on the predicted growth stage and nutritional indicators, and to generate harvest suggestions. The harvest reminder module is used to determine the optimal harvest time when the predicted growth stage is the nutrient enrichment period and the nutrient indicators reach the preset threshold, and to push a harvest reminder to the user.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images