Dendrobiaceae information system based on volatile organic compound fingerprint

Abstract

The application discloses a Dendrobiaceae popular science information system based on volatile organic compound (VOC) fingerprint, which mainly comprises a multi-channel VOC sensor array module integrated in a portable terminal, which is used for collecting VOCs released by dendrobium in real time and generating real-time VOC fingerprint data; a plant volatile fingerprint database module, which is used for storing standard VOC fingerprint vectors of different dendrobium species and physiological states; a volatile-physiological function mapping model module, which outputs the species, physiological state and key volatile interpretation of the dendrobium; a generative popular science content engine module, which dynamically generates personalized popular science texts based on a large language model and a user portrait; and an augmented reality interaction and visualization module, which superimposes a volatile concentration heat map, a physiological state label and an interactive popular science bubble on a real plant image. The application converts invisible chemical signals of the dendrobium into immersive popular science content, and significantly improves the interest and scientificity of popular science.

Description

Technical Field

[0001] This invention relates to the fields of smart agriculture and augmented reality technology, and more specifically, to a Dendrobium science popularization information system, display method, and electronic device based on volatile organic compound (VOC) fingerprint spectrum. Background Technology

[0002] Dendrobium spp. is a general term for plants of the genus Dendrobium in the family Orchidaceae. There are approximately 1500-1800 species worldwide, mainly distributed in subtropical regions south of the Yangtze River in my country, particularly concentrated in the southwest and south China. It possesses both high medicinal and ornamental value. Currently, plant science popularization systems based on mobile terminal image recognition (such as Baidu Image Search and Flower Identification) can provide users with species information, but these systems mostly provide static, homogenized text and image content, failing to perceive the plant's internal physiological state (such as stress response) or present the chemical interaction processes between the plant and its environment (such as allelopathy and volatile communication). International research also points out that existing apps have limited recognition accuracy in complex environments and insufficient reliability in high-risk scenarios. Therefore, the science popularization experience of existing systems has significant limitations in multimodal depth and biological information dimension. For example, patent CN113486205A aims to enhance 3D interaction through AR technology, but it focuses on technical solutions rather than a comprehensive improvement in user experience.

[0003] Volatile organic compounds (VOCs) are low-molecular-weight, high-vapor-pressure secondary metabolites released by plants, constituting the unique odor fingerprint of plants. Studies have shown that the VOCs of Dendrobium officinale encompass a variety of functional group compounds, including alkenes, alcohols, aldehydes, esters, ketones, phenols, and acids, and their composition and content are closely related to the physiological state of the plant. At the growth and development level, the types and contents of volatile components in Dendrobium chrysotoxum change systematically with the progress of flowering—gradually increasing from the bud stage to the early flowering stage, reaching a peak during the full bloom stage. The components released in the highest amounts during the full bloom stage are octyl acetate (accounting for 31.74% of the total content) and ocimene (24.98%), and then significantly decreasing towards the end of the flowering stage. Dendrobium primulinum exhibits a significant diurnal release rhythm, with the total aroma release accounting for 35.26% of the total daily release at 6:00 AM, decreasing to 7.17% by 6:00 PM, and the number of aroma components decreasing from 93 in the morning to 61 in the evening. Furthermore, Dendrobium devonianum is richest in terpenoids and volatile esters during its fully open period. Among the 664 VOCs detected, terpenoids such as α-humulene, β-bisabolene, and δ-juniperene are likely the main active components constituting its floral fragrance.

[0004] At the environmental response level, plants release specific stress-induced plant volatiles (HIPVs) when subjected to mechanical damage or insect infestation. Methyl jasmonic acid (MJA) is a key chemical messenger that transmits "danger signals" to neighboring plants. Studies have confirmed that treatment with exogenous MJA significantly alters both the metabolome and transcriptome of *Dendrobium nobile*, indicating that the jasmonic acid signaling pathway plays a central role in *Dendrobium nobile*'s defense response. When *Dendrobium nobile* suffers physical damage, the levels of volatiles such as MJA rapidly increase, acting as internal plant defense signals and interplant wireless communication mediators. Under normal physiological conditions, *Dendrobium nobile* continuously releases green leaf volatiles (GLVs), such as cis-3-hexenol, which has been detected in both the stems and leaves of *Dendrobium officinale*. Existing research has confirmed that GLVs inhibit the reproduction of microorganisms and various insects, suggesting their role in regulating the ecological balance of plant surface microorganisms and inhibiting pathogen colonization.

[0005] This demonstrates that the VOCs released by Dendrobium are not random chemical emissions, but rather "coupled signals" containing rich physiological information such as species identity, developmental stage, and environmental stress response. Existing research has constructed Dendrobium VOC fingerprint profiles based on ultrafast gas-phase electronic nose technology, and multivariate analysis methods can effectively distinguish different Dendrobium species and tissue parts. By detecting and analyzing the VOC fingerprint profiles of Dendrobium, it is possible to inversely infer the physiological state of the plant, thus providing a theoretical possibility and technical path for science popularization systems to break through the single dimension of image recognition, and is expected to open up new perceptual dimensions in areas such as visualization of plant physiological states and real-time monitoring of environmental responses.

[0006] However, there are no reports of deep integration between plant VOCs fingerprint detection and augmented reality science popularization systems in existing technologies. While publicly available gas sensing detection methods, such as CN120801438A, utilize electronic noses for rapid identification of crop pests, their applications are limited to pest diagnosis and fail to transform odor data into science education resources. Furthermore, existing AR science popularization systems lack input information on the plant's internal physiological state; their augmented reality content can only be statically and generically overlaid on the plant's appearance, unable to dynamically adjust the visualization theme and explanatory content based on the plant's real-time chemical information. A significant technological gap remains regarding the complete technology for real-time acquisition, intelligent identification, and dynamic science popularization generation of Dendrobium officinale VOCs fingerprint spectra. Summary of the Invention

[0007] The present invention aims to provide a Dendrobium science information system based on volatile organic compound fingerprinting to solve the technical problem that existing technologies cannot perceive the physiological state of plants in real time and generate personalized, immersive science content.

[0008] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0009] The Dendrobium science information system based on volatile organic compound fingerprinting mainly includes:

[0010] A multi-channel VOCs sensor array module, integrated into a portable terminal, contains several gas-sensitive elements that are sensitive to terpenes, aldehydes, esters, and jasmonic acid compounds, respectively. It is used to collect air around Dendrobium plants in real time, detect the composition and concentration of volatile organic compounds released by Dendrobium plants, and generate real-time VOCs fingerprint data.

[0011] The plant volatiles fingerprint database module is used to store standard VOCs fingerprints. The VOCs fingerprints are associated with at least the standard VOCs fingerprint vectors of different Dendrobium species and different physiological states and their corresponding physiological state labels.

[0012] The volatile matter-physiological function mapping model module uses a graph neural network. It takes the real-time fingerprint data collected by the multi-channel VOCs sensor array module as input, and performs feature matching and association reasoning with the standard VOCs fingerprint spectrum in the database module to output the current Dendrobium species identification, physiological state and key volatile matter interpretation.

[0013] The generative science popularization content engine module, based on a large language model, has its input connected to the output of the AI ​​volatiles-physiological function mapping model module and the user profile database, respectively. Based on the received plant status information and user profile, it calls preset science popularization content generation rules to dynamically generate personalized science popularization text.

[0014] The augmented reality interaction and visualization module is used to overlay and display a visualization layer corresponding to the plant status information on the real-time captured image of the mobile terminal device's camera; the layer includes: a volatile concentration heat map rendered on the outline of a real Dendrobium plant, physiological status labels floating above the plant, and interactive bubbles containing the personalized science popularization text.

[0015] Preferably, the target compounds detected by the multi-channel VOCs sensor array module described above include at least terpenes, aldehydes, esters, and jasmonic acids; the VOCs fingerprint data includes at least the sensor response values ​​corresponding to the terpenes, aldehydes, esters, and jasmonic acids.

[0016] Preferably, the physiological states described above include, but are not limited to, the full bloom period, the mechanical damage period, the normal growth period, and the early stage of disease infection.

[0017] Preferably, the volatile matter-physiological function mapping model module described above uses a graph neural network to embed features into real-time VOCs fingerprint data, generating corresponding high-dimensional volatile matter feature vectors; and uses a Mahalanobis distance discrimination algorithm based on the covariance matrix to perform state matching between the real-time VOCs feature vectors and the standard VOCs feature vectors in the plant volatile matter fingerprint database; wherein, by calculating the distribution distance between the real-time feature vectors and the standard feature vectors, the identification and reasoning of different Dendrobium species and physiological states are realized; when the Mahalanobis distance value is less than a preset threshold, it is determined that the current Dendrobium plant matches the state of the corresponding standard VOCs fingerprint spectrum.

[0018] Preferably, the generative science popularization content engine module described above stores a science popularization content generation rule base, which is used to match the corresponding science popularization text template according to the user profile, plant status and volatile types.

[0019] Preferably, the user profile information in the user profile database is obtained based on the registration information of the user's login terminal. The registration information includes at least an age tag or a knowledge level tag, which is used to classify users into user types including children, teenagers, adults, or professionals.

[0020] Preferably, in the augmented reality interaction and visualization module described above, the volatile concentration heatmap is color-coded and rendered in the corresponding organ region of the plant model according to the intensity of the response values ​​of each sensor.

[0021] A method for displaying Dendrobium science information system based on volatile organic compound fingerprinting, as described above, is characterized by the following steps:

[0022] (1) After the system starts, wait for the multi-channel VOCs sensor array module to warm up and stabilize for 60 seconds, and take the response readings of each sensor at this time as the current environmental background baseline R. i,0 Then, the air around the living Dendrobium plants is collected through the multi-channel VOCs sensor array module, and the data is processed according to formula S. i,j =(R i,j -R i,0 ) / R i,0 Calculate the real-time normalized response value of each sensor, apply exponential smoothing filter, and then use the instantaneous value at the current moment to construct the real-time VOCs fingerprint vector [S‾]. 1,j ,S‾ 2,j ,⋯ ,S‾ n,j ];

[0023] (2) Input the real-time VOCs fingerprint vector obtained in step (1) into the volatiles-physiological function mapping model, and compare it with the standard fingerprint in the plant volatiles fingerprint database module using the Mahalanobis distance dynamic matching algorithm based on graph neural network embedding features, and output the corresponding plant species and physiological state.

[0024] (3) Based on the registration information of the user login terminal, obtain the current user's profile information from the user profile database, wherein the profile information includes at least an age tag or a knowledge level tag;

[0025] (4) The generative popular science content engine module, based on the plant species and physiological state output in step (2) and the user profile obtained in step (3), calls the matching rules in the popular science content generation rule base to dynamically generate personalized popular science text;

[0026] (5) The augmented reality interaction and visualization module displays the physiological state output in step (2) as a label, the normalized response values ​​of each sensor in step (1) as a VOCs concentration heatmap, and the popular science text generated in step (4) as an interactive bubble on the real-time captured image of the plant in the AR view of the mobile terminal.

[0027] A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, it implements the steps of the above-described method for displaying Dendrobium officinale science popularization information system based on volatile organic compound fingerprinting.

[0028] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows:

[0029] 1. This invention breaks through the limitations of traditional plant science popularization systems that rely on image recognition. For the first time, it uses the fingerprint spectrum of plant volatile organic compounds (VOCs) as the sensing content. Through a multi-channel sensor array, it captures in real time the characteristic volatiles such as terpenes, aldehydes, esters, and jasmonic acids released by Dendrobium. It can inversely infer the physiological state of the plant, such as the flowering period, mechanical damage period, normal growth period, or early stage of disease infection, thus enriching the popular science content.

[0030] 2. This invention uses a portable terminal integrated VOCs sensor array, which does not require damaging plant tissue. Simply bring the sensor close to the plant to collect the gas, and fingerprint spectrum acquisition and identification can be completed. It is suitable for on-site environments such as botanical gardens, science education bases, and Chinese medicinal herb planting bases.

[0031] 3. This invention introduces a generative popular science content engine based on a large language model. It combines user login information (age tags, knowledge level tags) to distinguish users into children, teenagers, adults or professionals, and outputs popular science texts of different styles for the same plant physiological state, thus overcoming the shortcomings of homogeneous content in traditional popular science systems.

[0032] 4. This invention uses augmented reality technology to overlay the concentration of odor molecules, which is originally invisible, onto the outline of a real plant in the form of a heat map, allowing users to intuitively perceive the spatial distribution of volatile organic compounds; at the same time, the explanation is presented in an interactive bubble, which significantly enhances the intuitiveness and fun of popular science. Detailed Implementation

[0033] The present invention will be further described in detail below with reference to embodiments, so that those skilled in the art can implement it based on the description.

[0034] Example 1:

[0035] A Dendrobium science information system based on volatile organic compound fingerprinting mainly includes:

[0036] A multi-channel VOCs sensor array module, integrated into a portable terminal, contains several gas-sensitive elements that are sensitive to terpenes, aldehydes, esters, and jasmonic acid compounds, respectively. It is used to collect air around Dendrobium plants in real time, detect the composition and concentration of volatile organic compounds released by Dendrobium plants, and generate real-time VOCs fingerprint data.

[0037] The target compounds detected by the multi-channel VOCs sensor array module include at least terpenes, aldehydes, esters, and jasmonic acids; the VOCs fingerprint data includes at least the sensor response values ​​corresponding to the terpenes, aldehydes, esters, and jasmonic acids.

[0038] The plant volatile matter fingerprint database module is used to store standard VOC fingerprints. The VOC fingerprints are associated with at least the standard VOC fingerprint vectors of different Dendrobium species and under different physiological states, as well as their corresponding physiological state labels. The physiological states include, but are not limited to, the flowering period, the mechanical damage period, the normal growth period, and the early stage of disease infection.

[0039] The volatile matter-physiological function mapping model module uses a graph neural network. It takes the real-time fingerprint data collected by the multi-channel VOCs sensor array module as input, and performs feature matching and association reasoning with the standard VOCs fingerprint spectrum in the database module to output the current Dendrobium species identification, physiological state and key volatile matter interpretation.

[0040] The volatile matter-physiological function mapping model module uses a graph neural network to embed features into real-time VOCs fingerprint data, generating corresponding high-dimensional volatile matter feature vectors. It then uses a Mahalanobis distance discriminant algorithm based on the covariance matrix to perform state matching between the real-time VOCs feature vectors and the standard VOCs feature vectors in the plant volatile matter fingerprint database. Specifically, by calculating the distribution distance between the real-time feature vectors and the standard feature vectors, it achieves the identification and reasoning of different Dendrobium species and physiological states. When the Mahalanobis distance value is less than a preset threshold, it is determined that the current Dendrobium plant matches the corresponding standard VOCs fingerprint database.

[0041] The generative science popularization content engine module, based on a large language model, has its input connected to the output of the AI ​​volatiles-physiological function mapping model module and the user profile database, respectively. Based on the received plant status information and user profile, it calls preset science popularization content generation rules to dynamically generate personalized science popularization text.

[0042] The generative science popularization content engine module stores a science popularization content generation rule library, which is used to match the corresponding science popularization text template according to user profile, plant status and volatile types.

[0043] The user profile information in the user profile database is obtained based on the registration information of the user's login terminal. The registration information includes at least an age tag or a knowledge level tag, which is used to classify users into user types including children, teenagers, adults, or professionals.

[0044] The augmented reality interaction and visualization module is used to overlay and display a visualization layer corresponding to the plant status information on the real-time captured image of the mobile terminal device's camera; the layer includes: a volatile concentration heat map rendered on the outline of a real Dendrobium plant, physiological status labels floating above the plant, and interactive bubbles containing the personalized science popularization text.

[0045] In the augmented reality interaction and visualization module, the volatile concentration heatmap is color-coded and rendered in the corresponding organ region of the plant model based on the intensity of the response values ​​of each sensor.

[0046] In this embodiment, the user activates a portable terminal integrated with a multi-channel VOCs sensor array module and approaches the Dendrobium plant to be tested. The terminal collects air samples around the plant, and each sensor responds in real time, forming real-time VOCs fingerprint data. The portable terminal uploads the real-time VOCs fingerprint data to a cloud server via a wireless network. After receiving the data, the volatile matter-physiological function mapping model module compares it with the standard VOCs fingerprint spectrum in the plant volatile matter fingerprint spectrum database module, outputting the plant species, physiological state, and key volatile matter definitions. The generative science popularization content engine module receives the above output results and simultaneously obtains the current user's profile information (such as age tags and knowledge level tags) from the user profile database, matches it with science popularization text templates in the rule base, and dynamically generates personalized science popularization text. The augmented reality interaction and visualization module identifies the plant outline on the real-time image from the terminal's camera, overlays a heat map, physiological state tags, and interactive bubbles to complete the science popularization display.

[0047] Example 2:

[0048] This embodiment provides a method for displaying the Dendrobium science information system based on volatile organic compound fingerprinting as described in Embodiment 1, including the following steps:

[0049] Step (1): After the system starts, wait for the multi-channel VOCs sensor array module to warm up and stabilize for 60 seconds, and use the response readings of each sensor at this time as the current environmental background baseline R. i,0 Then, the air around the living Dendrobium plants is collected through the multi-channel VOCs sensor array module, and the data is processed according to formula S. i,j =(R i,j -R i,0 ) / R i,0 Calculate the real-time normalized response value of each sensor, apply exponential smoothing filter (smoothing coefficient is 0.3), and then take the instantaneous value at the current moment to construct the real-time VOCs fingerprint vector [S‾]. 1,j ,S‾ 2,j ,⋯ ,S‾ n,j ];

[0050] Step (2): Input the real-time VOCs fingerprint vector obtained in step (1) into the volatiles-physiological function mapping model, and compare it with the standard fingerprint in the plant volatiles fingerprint database module using the Mahalanobis distance dynamic matching algorithm based on graph neural network embedding features, and output the corresponding plant species and physiological state (such as flowering period, mechanical damage period, normal growth period, disease infection period, etc.).

[0051] Step (3): Based on the registration information of the user's login terminal, obtain the current user's profile information from the user profile database. The profile information includes at least an age tag or a knowledge level tag. Based on this, the system classifies the user into types such as children (6-12 years old), teenagers (13-18 years old), adults, or professionals.

[0052] Step (4): The generative science popularization content engine module, based on the plant species and physiological state output in step (2) and the user profile obtained in step (3), calls the matching rules in the science popularization content generation rule base; wherein the rule base stores the mapping relationship between "user profile - volatile species - plant state" and science popularization text template. This module fills the user profile, volatile names and plant states into the science popularization text template, and calls the large language model for polishing when necessary to generate the final personalized science popularization text, for example:

[0053] When the user profile is "preschool children (3-6 years old)," the key volatile compound is "ocimene," and the Dendrobium chrysanthum is in "peak flowering season," the matching science popularization text template is: "The Dendrobium flowers are blooming! Its fragrant 'scent signal' is saying: 'Buzz buzz buzz, little bees, come quick!'"

[0054] When the user profile is "children (7-12 years old)," the key volatile compound is "linalool," and the Dendrobium moniliforme is in "peak flowering," the matching educational text template is: "The Dendrobium is blooming with fragrant flowers; linalool is its 'scent name card,' telling butterflies and bees: 'I'm ready!'"

[0055] When the user profile is "teenagers (13-18 years old)," the key volatile substance is "methyl jasmonic acid," and the Dendrobium plant is in the "mechanical damage stage," the matching popular science text template is "When the leaves are bitten, Dendrobium releases methyl jasmonic acid as a 'distress signal' to tell its companions around it to activate defense."

[0056] When the user profile is "general public / gardening enthusiast", the key volatile compound is "leaf alcohol", and the Dendrobium plant is in "normal growth period", the matching popular science text template is "Can you smell a faint grassy scent from Dendrobium leaves? That's the scent of leaf alcohol. It's the basic background aroma released by healthy Dendrobium."

[0057] When the user profile is "Professional / Researcher", the key volatile substance is "Green Leaf Volatile (cis-3-Hexenol)", and the plant status is "Normal Growth Period", the matched popular science text template is: "High concentration of cis-3-Hexenol was detected. This is a typical green leaf volatile (GLV). Dendrobium maintains the microbial ecological balance on its surface by releasing these substances."

[0058] Step (5): The augmented reality interaction and visualization module displays the physiological state output in step (2) as a label, the normalized response values ​​of each sensor in step (1) as a VOCs concentration heatmap, and the popular science text generated in step (4) as an interactive bubble on the AR view of the mobile terminal, overlaying them onto the real-time captured image of the plant. Specifically: the outline of the Dendrobium plant is locked in real time by an image recognition algorithm (such as ORB feature point matching or deep learning target detection). According to the response intensity values ​​of each sensor in step (1), a volatile concentration heatmap is rendered in the corresponding organ area of ​​the plant: areas with high normalized response values ​​(such as flower buds) are displayed in red, areas with medium values ​​are displayed in yellow, and areas with low values ​​are displayed in green. Physiological state labels are displayed above the plant, such as "Dendrobium officinale | Peak flowering period". After the user clicks on the label, an interactive bubble containing the personalized popular science text generated in step (4) pops up.

[0059] Application Example 1:

[0060] User Xiaoming, 8 years old (automatically categorized as "child" by the system based on login information), brought his portable terminal close to a Dendrobium plant in full bloom. The terminal collected gas and generated real-time VOCs fingerprint data. After uploading to the server, the data was matched successfully if the Mahalanobis distance between the terminal and the standard fingerprint of "Dendrobium officinale_flowering period" in the database was the smallest and less than the system's preset threshold. The output of the volatile matter-physiological function mapping model is: Species = Dendrobium officinale, Physiological state = flowering period, Key volatile matter = linalool. Step (3) Obtain Xiaoming as a "child" type. The generative science content engine module matches the rule of "children + linalool + peak flowering period" to generate science text: "The Dendrobium is blooming! The linalool it releases is like sending a message to the bees: 'Come and collect honey!'" Then, Xiaoming raises his terminal screen and points it at the Dendrobium plant. The screen displays in real time: the flower bud area is presented as a red heat map, and the floating label "Dendrobium officinale | Peak Flowering Period" is displayed above. After clicking the label, a bubble pops up to display the above science text, accompanied by an AR animation of cartoon bees flying around.

[0061] Application Example 2:

[0062] User Zhang, a researcher (PhD in botany, automatically categorized as a "professional" by the system based on login information), was tested for the same Dendrobium species in use case 1. The volatile matter-physiological function mapping model output the same result: species = Dendrobium officinale, physiological state = peak flowering period, key volatile matter = linalool, but step (4) matched the rule "professional + linalool + peak flowering period", generating the popular science text: "High concentration of linalool was detected, which is a typical insect pollination signal, indicating that the plant is in the peak period of reproductive growth energy input." In the terminal screen view, the heat map also showed that the bud area was red, and the floating label "Dendrobium officinale | Peak flowering period" was displayed above. After clicking the label, the bubble displayed the above popular science text and the academic description about the expression of terpene synthase genes. It can be seen that the same system can provide completely differentiated popular science experiences according to user profiles.

[0063] Application Example 3:

[0064] User Xiao Li, 15 years old (automatically categorized as a "teenager" based on login information), brought his portable terminal close to a Dendrobium nobile plant with damaged leaves. The terminal collected gas and generated real-time VOCs fingerprint data. The response value of the jasmonic acid sensor increased significantly. The Mahalanobis distance between the sensor and the standard fingerprint "Dendrobium nobile_mechanical damage period" in the database was the smallest and less than the system's preset threshold. Therefore, the match was successful. The volatile matter-physiological function mapping model output: Species = Dendrobium nobile, Physiological state = mechanical damage period, Key volatile matter = methyl jasmonic acid. Step (3) identifies Xiao Li as a "teenager". The generative science content engine module matches the rule "teenager + methyl jasmonate + mechanical damage period" to generate the science text: "When the leaves are bitten, Dendrobium releases methyl jasmonate as a 'distress signal' to tell its surrounding companions to activate defense." The terminal screen view renders an orange heat map in the damaged leaf area and dynamically flashes the "SOS" label. Clicking the label displays the above science text in a bubble, accompanied by an animation illustrating the signal transmission between plant cells. Users can understand the principles of plant defense ecology through interaction.

[0065] Example 3:

[0066] This embodiment provides a computer-readable storage medium, which may be ROM, RAM, disk, USB flash drive, SD card, or cloud virtual storage space, etc. The storage medium stores a computer program, which, when executed by a processor (such as the CPU of a portable terminal), implements all the steps of the method shown in Embodiment 2.

[0067] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A Dendrobium science popularization information system based on volatile organic compound fingerprinting, characterized in that: Mainly includes: A multi-channel VOCs sensor array module, integrated into a portable terminal, contains several gas-sensitive elements that are sensitive to terpenes, aldehydes, esters, and jasmonic acid compounds, respectively. It is used to collect air around Dendrobium plants in real time, detect the composition and concentration of volatile organic compounds released by Dendrobium plants, and generate real-time VOCs fingerprint data. The plant volatiles fingerprint database module is used to store standard VOCs fingerprints. The VOCs fingerprints are associated with at least the standard VOCs fingerprint vectors of different Dendrobium species and different physiological states and their corresponding physiological state labels. The volatile matter-physiological function mapping model module uses a graph neural network. It takes the real-time fingerprint data collected by the multi-channel VOCs sensor array module as input, and performs feature matching and association reasoning with the standard VOCs fingerprint spectrum in the database module to output the current Dendrobium species identification, physiological state and key volatile matter interpretation. The generative science popularization content engine module is based on a large language model. Its input is connected to the output of the volatile-physiological function mapping model module and the user profile database. Based on the received plant status information and user profile, it calls the preset science popularization content generation rules to dynamically generate personalized science popularization text. The augmented reality interaction and visualization module is used to overlay and display a visualization layer corresponding to the plant status information on the real-time captured image of the mobile terminal device's camera; the layer includes: a volatile concentration heat map rendered on the outline of a real Dendrobium plant, physiological status labels floating above the plant, and interactive bubbles containing the personalized science popularization text.

2. The Dendrobium science popularization information system based on volatile organic compound fingerprinting according to claim 1, characterized in that: The target compounds detected by the multi-channel VOCs sensor array module include at least terpenes, aldehydes, esters, and jasmonic acids; the VOCs fingerprint data includes at least the sensor response values ​​corresponding to the terpenes, aldehydes, esters, and jasmonic acids.

3. The Dendrobium science popularization information system based on volatile organic compound fingerprinting according to claim 1, characterized in that: The physiological states mentioned include, but are not limited to, the flowering period, the mechanical damage period, the normal growth period, and the early stage of disease infection.

4. The Dendrobium science popularization information system based on volatile organic compound fingerprinting according to claim 1, characterized in that: The volatile matter-physiological function mapping model module uses a graph neural network to embed features into real-time VOCs fingerprint data, generating corresponding high-dimensional volatile matter feature vectors. It then uses a Mahalanobis distance discriminant algorithm based on the covariance matrix to perform state matching between the real-time VOCs feature vectors and the standard VOCs feature vectors in the plant volatile matter fingerprint database. Specifically, by calculating the distribution distance between the real-time feature vectors and the standard feature vectors, it achieves the identification and reasoning of different Dendrobium species and physiological states. When the Mahalanobis distance value is less than a preset threshold, it is determined that the current Dendrobium plant matches the corresponding standard VOCs fingerprint database.

5. The Dendrobium science popularization information system based on volatile organic compound fingerprinting according to claim 1, characterized in that: The generative science popularization content engine module stores a science popularization content generation rule base, which is used to match corresponding science popularization text templates based on user profiles, plant conditions, and types of volatile substances.

6. The Dendrobium science popularization information system based on volatile organic compound fingerprinting according to claim 1, characterized in that: The user profile information in the user profile database is obtained based on the registration information of the user's login terminal. The registration information includes at least an age tag or a knowledge level tag, which is used to classify users into user types including children, teenagers, adults, or professionals.

7. The Dendrobium science popularization information system based on volatile organic compound fingerprinting according to claim 1, characterized in that: In the augmented reality interaction and visualization module, the volatile concentration heatmap is color-coded and rendered in the corresponding organ region of the plant model based on the intensity of the response values ​​of each sensor.

8. A method for displaying Dendrobium science information system based on volatile organic compound fingerprinting as described in any one of claims 1 to 7, characterized in that: Includes the following steps: (1) After the system starts, wait for the multi-channel VOCs sensor array module to warm up and stabilize for 60 seconds, and take the response readings of each sensor at this time as the current environmental background baseline R. i,0 Then, the air around the living Dendrobium plants is collected through the multi-channel VOCs sensor array module, and the data is processed according to formula S. i,j =(R i,j -R i,0 ) / R i,0 Calculate the real-time normalized response value of each sensor, apply exponential smoothing filter, and then use the instantaneous value at the current moment to construct the real-time VOCs fingerprint vector [S‾]. 1,j ,S‾ 2,j ,⋯ ,S‾ n,j ]; (2) Input the real-time VOCs fingerprint vector obtained in step (1) into the volatiles-physiological function mapping model, and compare it with the standard fingerprint in the plant volatiles fingerprint database module using the Mahalanobis distance dynamic matching algorithm based on graph neural network embedding features, and output the corresponding plant species and physiological state. (3) Based on the registration information of the user login terminal, obtain the current user's profile information from the user profile database, wherein the profile information includes at least an age tag or a knowledge level tag; (4) The generative popular science content engine module, based on the plant species and physiological state output in step (2) and the user profile obtained in step (3), calls the matching rules in the popular science content generation rule base to dynamically generate personalized popular science text; (5) The augmented reality interaction and visualization module displays the physiological state output in step (2) as a label, the normalized response values ​​of each sensor in step (1) as a VOCs concentration heatmap, and the popular science text generated in step (4) as an interactive bubble on the real-time captured image of the plant in the AR view of the mobile terminal.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the display method as described in claim 8.