Terahertz metamaterial sensor array-based crop disease detection method and detection device

Abstract

The application provides a crop disease detection method and device based on a terahertz metamaterial sensor array, and comprises the following detection steps: S1: using a detachable enrichment unit to adsorb target objects on the surface of a living leaf of a crop to be detected, wherein the target objects include pathogenic microorganism spores and volatile organic compounds; under the condition of 15-30 DEG C, the enrichment time is 2-3 minutes; under the condition of less than 15 DEG C, the enrichment time is 3-5 minutes; the application combines a functionalized biological probe modified heterogeneous terahertz metamaterial sensor array with an AI model, realizes accurate capture of early trace markers of diseases by virtue of the local field enhancement effect of the metamaterial, and effectively offsets environmental interference through a reference unit; the device is designed in a handheld integrated manner, can be used for in-situ detection in a field without damaging crops, is simple to operate, and directly solves the pain points of difficulty in early detection, large interference and poor practicability of traditional methods.

Description

Technical Field

[0001] This invention relates to the field of crop disease detection technology, specifically to a crop disease detection method and device based on a terahertz metamaterial sensor array. Background Technology

[0002] Early and accurate diagnosis of crop diseases is a core aspect of agricultural disaster prevention and mitigation. The key lies in the effective capture and identification of trace markers (such as pathogenic spores and volatile organic compounds) released in the early stages of disease infection. Terahertz waves, as electromagnetic waves between microwaves and infrared radiation, possess a unique "fingerprint spectrum" recognition capability for the low-frequency rotation and vibration modes of biological macromolecules, distinguishing the differences in biochemical composition between healthy and diseased crops. Metamaterials, through artificially designed subwavelength resonant structures, can generate extremely strong local electromagnetic field enhancement effects, elevating the sensitivity of terahertz sensing to the level of trace detection. The combination of these two technologies provides core technological support for overcoming the bottleneck in early detection of crop diseases.

[0003] The shortcomings of current mainstream detection technologies have become key obstacles to the control of diseases in the field: visual observation relies on the experience of agricultural technicians and can only identify obvious symptoms such as yellowing and necrotic spots on leaves. It is completely unable to determine the incubation period of diseases (when the pathogen has infected but has not yet shown symptoms), resulting in a delay in the timing of control; molecular detection such as PCR and ELISA has high accuracy, but the samples need to undergo complex pretreatment such as grinding, reagent extraction, and isothermal amplification, which not only damages crop tissue, but also requires the entire process to be completed in the laboratory. It takes several hours from sample collection to results, and it is completely unable to detect in situ in the field; optical imaging technologies such as near-infrared and hyperspectral imaging achieve non-destructive detection, but their detection principle relies on changes in the optical reflectance characteristics of the crop surface. However, the small fluctuations (usually less than 5%) of the internal biochemical components of crops in the early stage of disease (such as chlorophyll and enzyme activity) cannot cause significant differences in optical signals, resulting in insufficient sensitivity. Moreover, it is easily interfered with by environmental factors such as light intensity, leaf water content, and field dust, resulting in a high misjudgment rate. Even in recent years, the research on terahertz metamaterial sensing has focused on the performance verification of single sensors in the laboratory. These sensors can only detect single disease indicators and lack correction mechanisms for field temperature and humidity and interference from non-target molecules. Moreover, the devices are mostly large desktop equipment with low integration and complex operation, which cannot meet the needs of grassroots agricultural technicians for portable use in the field. Summary of the Invention

[0004] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a method and device for detecting crop diseases based on terahertz metamaterial sensor arrays, which solves the problem that the deficiencies of mainstream detection technologies have become a key obstacle to the prevention and control of diseases in the field.

[0005] (II) Technical Solution To achieve the above objectives, the present invention provides a method for detecting crop diseases based on a terahertz metamaterial sensor array, comprising the following detection steps: S1: A detachable enrichment unit is used to adsorb target substances from the surface of living leaves of the crop to be tested. The target substances include pathogenic microbial spores and volatile organic compounds. The enrichment time is 2 to 3 minutes at 15 to 30 degrees Celsius and 3 to 5 minutes at temperatures below 15 degrees Celsius. S2: Place the enrichment unit containing the target object in the detection area of ​​the terahertz metamaterial sensing array, scan the sensing array using a terahertz source, and receive the response signal of the sensing array through a detector. S3: Perform baseline correction and filtering on the response signal in sequence, extract the resonant peak frequency offset and intensity change between the detection unit and the reference unit in the sensor array, and calculate the effective frequency offset using a preset formula to correct for environmental interference. S4: Input the feature vector containing the effective frequency offset and intensity change into the pre-trained disease diagnosis model, and output the disease diagnosis results, which include the disease type, confidence level and severity level.

[0006] Preferably, the detection limit of the target analyte is: volatile organic compounds not exceeding 2 ppb.

[0007] Preferably, the terahertz metamaterial sensing array is a heterogeneous array, including a detection unit, a reference unit, and a calibration unit; the surface of the detection unit is modified with a specific biological probe, the surface of the reference unit is modified with a non-specific blocking agent, and the surface of the calibration unit is modified with a standard marker.

[0008] A crop disease detection device based on terahertz metamaterial sensor array includes a handheld detection probe, a gas enrichment unit, a main control and analysis module, and a power supply module. The handheld detection probe is used to emit terahertz waves, receive response signals, and complete preliminary signal acquisition. The gas enrichment unit is detachably connected to the handheld detection probe and is used to adsorb and enrich the target substances on the surface of the crop leaves to be tested. The main control and analysis module is electrically connected to the handheld detection probe and is used to control the detection process, process the collected signals, and run the disease diagnosis model. The power module is electrically connected to the handheld detection probe, the gas enrichment unit, and the main control and analysis module, respectively, and is used to provide working power.

[0009] Preferably, the handheld detection probe includes a 3-5THz quantum cascade laser, a microbolometer, and a heterogeneous sensor array.

[0010] Preferably, the gas enrichment unit includes a polytetrafluoroethylene shell, a replaceable activated carbon-PDMS composite enrichment membrane, and a micro fan.

[0011] Preferably, the main control and analysis module includes an STM32H743 processor, an eMMC storage unit, and a LoRa / USB communication unit.

[0012] Preferably, the power module is a 5000mAh rechargeable lithium battery; the detection unit of the sensing array includes an open-loop resonator, the surface of which is coated with a 50-100nm gold film and a specific probe is fixed by silanization.

[0013] (III) Beneficial Effects This invention provides a method and device for detecting crop diseases based on a terahertz metamaterial sensor array. It has the following beneficial effects: 1. This invention combines a heterogeneous terahertz metamaterial sensing array modified with functionalized biological probes with an AI model. It achieves accurate capture of trace markers in the early stages of disease by leveraging the local field enhancement effect of metamaterials. The reference unit effectively counteracts environmental interference. The device features a handheld design, allowing for in-situ field detection without damaging crops. It is simple to operate and directly addresses the pain points of traditional methods, such as difficulty in early detection, significant interference, and poor practicality. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall structure of the crop disease detection method and device based on terahertz metamaterial sensor array proposed in this invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: like Figure 1 As shown, this embodiment of the invention provides a crop disease detection method based on a terahertz metamaterial sensor array, including the following detection steps: S1: A detachable enrichment unit is used to adsorb target substances from the surface of living leaves of the crop to be tested. The target substances include pathogenic microbial spores and volatile organic compounds. The enrichment time is 2 to 3 minutes at 15 to 30 degrees Celsius and 3 to 5 minutes at temperatures below 15 degrees Celsius. S2: Place the enrichment unit containing the target object in the detection area of ​​the terahertz metamaterial sensing array, scan the sensing array using a terahertz source, and receive the response signal of the sensing array through a detector. S3: Perform baseline correction and filtering on the response signal in sequence, extract the resonant peak frequency offset and intensity change between the detection unit and the reference unit in the sensor array, and calculate the effective frequency offset using a preset formula to correct for environmental interference. S4: Input the feature vector containing the effective frequency offset and intensity change into the pre-trained disease diagnosis model, and output the disease diagnosis results, including the disease type, confidence level and severity level.

[0017] The detection limit for the target analyte is: no more than 2 ppb for volatile organic compounds.

[0018] The terahertz metamaterial sensing array is a heterogeneous array, including a detection unit, a reference unit, and a calibration unit. The surface of the detection unit is modified with a specific biological probe, the surface of the reference unit is modified with a non-specific blocking agent, and the surface of the calibration unit is modified with a standard marker.

[0019] The feasibility and effectiveness of the method were verified using early detection of wheat stripe rust as an application scenario. 1. Core Preparation 1.1 Heterogeneous sensor array enabled: The heterogeneous terahertz metamaterial sensing array of this invention is selected, comprising 3 detection units (surface-modified wheat stripe rust IgY specific probe), 1 reference unit (modified with BSA blocking agent), and 1 calibration unit (modified with 10 ng / mL stripe rust standard protein). Before use, calibration is performed using the calibration unit to ensure that the resonance peak shift deviation is ≤5%.

[0020] 1.2 Loading the diagnostic model: Load the pre-trained SVM model, which is trained based on the feature vectors of 200 healthy samples, 180 latent samples, and 150 early stripe rust samples, with a validation set accuracy of 95%.

[0021] 2. Specific steps for implementation: 2.1 S1: Sample enrichment Operation method: Place the detachable enrichment unit close to the surface of the wheat leaf to be tested, and start the micro fan inside the unit.

[0022] Key parameters implemented: The actual field temperature was 28°C. The enrichment time was 2 minutes under the conditions of 15 to 30°C according to the method of this invention. If the temperature is 12°C, the enrichment time is extended to 4 minutes.

[0023] Target analyte capture: An enrichment membrane (activated carbon-PDMS composite membrane) adsorbs stripe rust spores from the leaf surface (concentration approximately 3 spores / mm). 2 The concentration of spores and volatile organic compounds (VOCs, approximately 1.5 ppb) meets the detection limit requirement set by this invention: "spores ≤ 5 / mm²". 2 The detection limit coverage is "VOCs≤2ppb".

[0024] 2.2 S2: Terahertz Detection Equipment docking: Secure the enrichment unit to the detection area of ​​the terahertz detection module, ensuring that the enrichment membrane and the sensor array are aligned.

[0025] Signal acquisition: Start the 3.5THz quantum cascade laser scanning array, scanning range 3-4THz; receive the reflective response signal through a microbolometer, continuously acquire for 3 seconds, and store the original time domain signal.

[0026] 2.3 S3: Signal Processing Preprocessing: The time-domain signal is subjected to third-order polynomial baseline correction + 5nm window Gaussian filtering to remove noise and drift.

[0027] Feature extraction and interference correction: The frequency domain spectrum is converted by FFT, and the resonant peak shift ΔF1 (0.03THz) and intensity change ΔI1 (12%) of the detection unit are extracted; at the same time, ΔF2 (0.008THz) of the reference unit is extracted. The effective ΔF after correction is calculated according to the formula "effective ΔF=ΔF1-ΔF2×0.03" (0.03 is the calibration environment coefficient), forming the feature vector [0.0298THz, 12%].

[0028] 2.4 S4: Intelligent Recognition Model inference: Input the feature vector into the SVM model and output the diagnostic results: "Disease type: wheat stripe rust; confidence level: 94%; severity level: level 1 (latency period)".

[0029] Output result: Simultaneously generated prevention and control recommendation: "It is recommended to spray 25% cyazofamid EC at a dilution of 1500 times within 7 days".

[0030] 3. Method effectiveness verification Detection limit validation: Using a stripe rust VOCs standard gas with a concentration of 2 ppb and 5 particles / mm 2 All spore samples were consistently detected (detection rate 100%), meeting the detection limit requirements set by this invention.

[0031] Anti-interference verification: Under high humidity conditions of 35℃ / 80%RH, after calibration by the reference unit, the false judgment rate is only 3%, which meets the detection limit requirements set by this invention.

[0032] Early validation: The detection rate of wheat 3 days after inoculation with stripe rust (asymptomatic) was 92%, which is significantly better than visual observation (detection rate of 0%) and near-infrared detection (detection rate of 40%).

[0033] Example 2: A crop disease detection device based on terahertz metamaterial sensor array includes a handheld detection probe, a gas enrichment unit, a main control and analysis module, and a power supply module. The handheld detection probe is used to emit terahertz waves, receive response signals, and complete preliminary signal acquisition. The gas enrichment unit is detachably connected to the handheld detection probe and is used to adsorb and enrich the target substances on the surface of the crop leaves to be tested. The main control and analysis module is electrically connected to the handheld detection probe and is used to control the detection process, process the collected signals, and run the disease diagnosis model. The power module is electrically connected to the handheld detection probe, the gas enrichment unit, and the main control and analysis module to provide operating power.

[0034] The handheld detection probe includes a 3-5THz quantum cascade laser, a microbolometer, and a heterogeneous sensor array.

[0035] The gas enrichment unit includes a polytetrafluoroethylene shell, a replaceable activated carbon-PDMS composite enrichment membrane, and a micro fan.

[0036] The main control and analysis module includes an STM32H743 processor, an eMMC storage unit, and a LoRa / USB communication unit.

[0037] The power module is a 5000mAh rechargeable lithium battery; the detection unit of the sensing array contains an open-loop resonator, and the surface is covered with a 50-100nm gold film and a specific probe is fixed by silanization.

[0038] 1. Core Component Structure and Implementation 1.1 Handheld detection probe Structural composition: ABS plastic outer shell, internal integration: Terahertz source: 3-5THz quantum cascade laser, fixed in a heat sink; Detector: 320×240 pixel microbolometer, 1cm spacing from the sensor array; Sensor array: 2×2mm heterogeneous array, fixed to the detection area by a metal bracket, with a reserved interface for enrichment unit snap-fit.

[0039] Core functions: Terahertz wave transmission, signal reception and preliminary acquisition are completed within 10 seconds, with optical path reflectivity ≥95%.

[0040] 1.2 Gas Enrichment Unit Structural composition: Polytetrafluoroethylene outer shell, internal structure: Replaceable enrichment membrane: activated carbon-PDMS composite membrane, fixed by a slot; Miniature fan: 3.7V DC brushless fan, wires connect to the power interface.

[0041] Usage: The snap-on probe allows for quick replacement of the enrichment membrane after each test, preventing cross-contamination.

[0042] 1.3 Main Control and Analysis Module Structure: Based on the STM32H743 microprocessor, it integrates: Storage unit: 16GB eMMC chip (stores raw data and SVM model); Communication unit: LoRa module + USB Type-C interface; Display unit: 2.4-inch LCD screen.

[0043] Core functions: Complete signal processing and model inference within 0.5 seconds, and support local display and remote uploading of detection results.

[0044] 1.4 Power Module Structural parameters: 5000mAh 3.7V rechargeable lithium battery with overcharge / overdischarge protection board and Type-C charging interface.

[0045] Performance: After 100 consecutive tests, the battery level remained at 18%, and the standby time was 50 hours, meeting the needs of all-day field operations.

[0046] 2. Device Integration and Field Application 2.1 Integration Method Each component is assembled as an integrated unit: the main control module is embedded in the probe handle, the power module is placed at the tail of the handle, and the enrichment unit is connected to the front of the probe via a snap-fit, which meets the requirements of handheld operation.

[0047] 2.2 Verification of Usage Effect Convenience: Grassroots agricultural technicians can operate independently after 15 minutes of training, and the testing time for a single wheat plant is ≤3 minutes.

[0048] Accuracy: Field testing of 100 wheat plants showed 94% consistency with laboratory ELISA results, with a 91% detection rate of latent diseases.

[0049] Environmental interference resistance: Within the range of 10-35℃ and 40%-80%RH, after calibration by the reference unit, the misjudgment rate is ≤4%.

[0050] Multi-scenario adaptation: In addition to wheat stripe rust, by changing the probe of the detection unit, it can detect rice blast, corn leaf blight, wheat powdery mildew and wheat scab, realizing "one device for multiple disease detection".

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for detecting crop diseases based on terahertz metamaterial sensor arrays, characterized in that, The following testing steps are included: S1: A detachable enrichment unit is used to adsorb target substances from the surface of living leaves of the crop to be tested. The target substances include pathogenic microbial spores and volatile organic compounds. The enrichment time is 2 to 3 minutes at 15 to 30 degrees Celsius and 3 to 5 minutes at temperatures below 15 degrees Celsius. S2: The enrichment unit containing the target substance is placed in the detection area of ​​the terahertz metamaterial sensing array. The sensing array is scanned using a terahertz source, and the response signal of the sensing array is received by a detector. The terahertz metamaterial sensing array is a heterogeneous array, including a detection unit, a reference unit, and a calibration unit. The surface of the detection unit is modified with a specific biological probe, the surface of the reference unit is modified with a non-specific blocking agent, and the surface of the calibration unit is modified with a standard marker. S3: Perform baseline correction and filtering on the response signal in sequence, extract the resonant peak frequency offset and intensity change between the detection unit and the reference unit in the sensor array, and calculate the effective frequency offset using a preset formula to correct for environmental interference. S4: Input the feature vector containing the effective frequency offset and intensity change into the pre-trained disease diagnosis model, and output the disease diagnosis results, which include the disease type, confidence level and severity level.

2. The crop disease detection method based on terahertz metamaterial sensor array according to claim 1, characterized in that: The detection limit for the target analyte is: volatile organic compounds not exceeding 2 ppb.

Images