Kiwi fruit disease identification method, device and equipment based on hyperspectrum and medium

Abstract

The invention provides a hyperspectrum-based kiwi fruit disease identification method, device and equipment and a medium. The hyperspectrum-based kiwi fruit disease identification method comprises the following steps: acquiring and correcting a hyperspectral image of a to-be-detected kiwi fruit sample; performing region-of-interest segmentation and spectral data extraction on the corrected hyperspectral image; preprocessing the extracted spectral data; screening characteristic wavelengths of the preprocessed spectral data through a genetic algorithm; and completing kiwi fruit disease identification on the screened characteristic wavelengths through a constructed disease identification model. The method has the advantages of simplicity, wide range, no damage and the like, the occurrence degree of diseases can be accurately found as early as possible, corresponding prevention and control measures can be taken, the loss of fruit farmers is effectively reduced, and economic benefits are increased.

Description

Technical Field

[0001] The invention relates to a kiwi disease identification method, device, equipment and medium based on hyperspectral, and relates to the field of plant disease identification. Background Art

[0002] Kiwifruit (Actinidia chinensis) is a deciduous fruit tree of the genus Actinidia. It is widely popular among the public because its fruit is rich in nutrients and vitamin C.

[0003] However, changes in climate, environment and production methods have led to increasingly serious kiwifruit diseases, which have seriously affected the yield and quality of kiwifruit and restricted the development of the industry. Among them, kiwifruit storage diseases caused by Botryosphaeria dothidea are common in its planting areas and are a serious fungal disease during storage.

[0004] Identification of kiwifruit diseases during storage is the basis of disease prevention and control. Traditional disease identification is mainly through indoor pathogen culture and manual identification, which is not only labor-intensive and material-intensive, but also affects the accuracy of prediction to a certain extent. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, in view of the above problems, the purpose of the present invention is to provide a kiwifruit disease identification method, device, equipment and medium based on hyperspectral that can accurately identify fungal diseases during kiwifruit storage.

[0006] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is:

[0007] In a first aspect, the present invention provides a kiwifruit disease identification method based on hyperspectral, the method comprising:

[0008] Acquire and calibrate the hyperspectral images of the kiwifruit samples to be tested;

[0009] Segment the region of interest and extract spectral data from the corrected hyperspectral image;

[0010] Preprocessing the extracted spectral data;

[0011] The preprocessed spectral data is used to select characteristic wavelengths through a genetic algorithm;

[0012] The selected characteristic wavelengths are used to complete kiwifruit disease identification through the constructed disease identification model.

[0013] In a possible implementation, the hyperspectral image is collected and corrected using FIGSPEC FS10 series hyperspectral imager and the matching image collection software FigSpec.

[0014] In one possible implementation, the corrected hyperspectral image is segmented into regions of interest and spectral data is extracted. The specific process is as follows: a sample region of interest ROI is extracted from the acquired hyperspectral image, and the average spectral information of the region of interest is used as the original spectral information, which is subjected to spectral preprocessing to remove the noise bands at both ends.

[0015] In a possible implementation manner, the extracted spectral data is preprocessed using standard normal variable transformation, multivariate scattering correction, SG convolution smoothing and first-order derivative to preprocess the spectral data to reduce noise caused by various factors.

[0016] In a possible implementation, the construction of the disease identification model includes:

[0017] preparing kiwi fruit samples;

[0018] The prepared kiwifruit samples were subjected to hyperspectral image acquisition and correction, region of interest segmentation and spectral data extraction, spectral data preprocessing and characteristic wavelength screening to obtain experimental samples;

[0019] The KS algorithm is used to divide the experimental samples into training sets and test sets in proportion;

[0020] The training set is used to train the disease recognition model based on the selected classifier, and the disease recognition model is verified during the training process through the test set.

[0021] In a possible implementation, the classifier uses a random forest, a support vector machine, a back propagation neural network, or an extreme learning machine.

[0022] In a possible implementation manner, preparing a kiwi fruit sample comprises:

[0023] After cleaning and sterilization, the kiwifruit samples were kept at room temperature of 20°C and relative humidity of RH = 45% for a set time.

[0024] The selected kiwifruit fruits were inoculated with B. dothidea strains, and a control group without inoculation was set up at the same time;

[0025] After inoculation, all kiwifruits were moisturized and stored with the inoculated surface facing upward at room temperature of 25°C.

[0026] The day of inoculation was recorded as day 0, and hyperspectral images were collected when the set number of days after inoculation was met.

[0027] In a second aspect, the present invention further provides a kiwifruit disease identification device based on hyperspectral, the device comprising:

[0028] A hyperspectral data acquisition unit is configured to acquire and correct a hyperspectral image of a kiwifruit sample to be tested;

[0029] A data segmentation and extraction unit is configured to segment the region of interest and extract spectral data from the corrected hyperspectral image;

[0030] A preprocessing unit, configured to preprocess the extracted spectral data;

[0031] A wavelength screening unit is configured to screen characteristic wavelengths from the preprocessed spectral data through a genetic algorithm;

[0032] The disease identification unit is configured to complete kiwifruit disease identification by using the selected characteristic wavelengths through the constructed disease identification model.

[0033] In a third aspect, the present invention also provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor so that the processor can execute any one of the methods described.

[0034] In a fourth aspect, the present invention further provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include computer instructions, and the computer instructions are used to enable a computer to execute any one of the methods described.

[0035] The present invention adopts the above technical solution, and has the following characteristics:

[0036] 1. Hyperspectral technology is an imaging data technology based on multiple narrow bands. The present invention uses the electromagnetic spectrum to study the spectral characteristics of various substances in the form of images for the identification and monitoring of plant diseases. This technology is used to identify fungal diseases of kiwifruit during storage. It has the advantages of simplicity, wide range, and no damage. It can detect the extent of disease occurrence early and accurately and take corresponding prevention and control measures, effectively reducing the losses of fruit farmers and increasing economic benefits.

[0037] 2. In view of the problem that fungal diseases of existing kiwifruit occur seriously during storage and are difficult to identify at an early stage, the hyperspectral-based kiwifruit disease identification method provided by the present invention utilizes different preprocessing and modeling methods to perform modeling and prediction.

[0038] In summary, the present invention can be widely used in the identification of plant diseases such as kiwifruit. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Throughout the accompanying drawings, the same reference numerals are used to represent the same components. In the accompanying drawings:

[0040] Figure 1 The kiwi fruit sample according to the embodiment of the present invention;

[0041] Figure 2 This is a hyperspectral image acquisition system according to an embodiment of the present invention, wherein FIG. (a) is an experimental framework, and FIG. (b) is a real object effect diagram.

[0042] Figure 3 : The segmentation of the region of interest in the embodiment of the present invention, wherein (a) is the ROI of fresh kiwifruit, (b) is the ROI of inoculated kiwifruit, and (c) is the ROI of naturally diseased kiwifruit;

[0043] Figure 4 is the average spectrum information of the embodiment of the present invention;

[0044] Figure 5 It is a comparison diagram of the pretreatment of the healthy sample and the Botryosphaeria dothidea sample of the embodiment of the present invention, wherein (a) is the spectrum after RAW pretreatment, (b) is the spectrum after SNV pretreatment, (c) is the spectrum after MSC pretreatment, (d) is the spectrum after SG pretreatment, and (e) is the spectrum after FD pretreatment;

[0045] Figure 6 The accuracy of different preprocessing methods and modeling methods in the embodiments of the present invention;

[0046] Figure 7 FIG. 4 is a structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0048] Although the terms first, second, third, etc. can be used in the text to describe multiple elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms can only be used to distinguish an element, component, region, layer or section from another region, layer or section. Unless the context clearly indicates, terms such as "first", "second" and other numerical terms do not imply order or sequence when used in the text. Therefore, the first element, component, region, layer or section discussed below can be referred to as the second element, component, region, layer or section without departing from the teaching of the example embodiments.

[0049] For ease of description, spatially relative terms may be used herein to describe the relationship of one element or feature relative to another element or feature as shown in the figures, such as "inside", "outside", "inner side", "outer side", "below", "above", etc. Such spatially relative terms are intended to include different orientations of the device in use or operation in addition to the orientation depicted in the figures.

[0050] Traditional disease identification is mainly through indoor pathogen cultivation and manual identification. This identification method is not only labor-intensive and material-intensive, but also affects the accuracy of prediction to a certain extent. The hyperspectral-based kiwifruit disease identification method, device, equipment and medium provided by the present invention include: collecting and correcting hyperspectral images of kiwifruit samples to be detected; segmenting the region of interest and extracting spectral data of the corrected hyperspectral images; preprocessing the extracted spectral data; screening characteristic wavelengths of the preprocessed spectral data through a genetic algorithm; and completing kiwifruit disease identification through the constructed disease identification model using the screened characteristic wavelengths. Therefore, the present invention can accurately discover the extent of disease occurrence and take corresponding prevention and control measures, effectively reducing the losses of fruit farmers and increasing economic benefits.

[0051] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0052] Embodiment 1: The hyperspectral-based kiwifruit disease identification method provided in this embodiment includes:

[0053] S1. Sample preparation

[0054] In this embodiment, the process of sample preparation is:

[0055] S11. After uniform cleaning and sterilization, the kiwifruit experimental samples were kept at room temperature of 20°C and relative humidity of RH=45% for 24 hours. The purpose was to eliminate the pathogens in the samples and prevent other pathogens from affecting the experimental results. At the same time, the physical and chemical properties of the kiwifruit were stabilized and the influence of temperature on data collection was eliminated.

[0056] S12. The selected kiwifruit fruits are used to inoculate the B. dothidea strain, and a control group without inoculation is set up at the same time.

[0057] S13. After the inoculation is completed, all kiwifruits are moisturized and stored at room temperature at 25°C with the inoculated surface facing upward. The day of inoculation is recorded as day 0, and hyperspectral images are collected 7 days after inoculation (this is an example, not limited to this). Figure 1 Shown is a kiwi fruit sample.

[0058] S2. Acquisition and correction of hyperspectral images

[0059] In this embodiment, FIGSPEC FS10 series hyperspectral imager is used to collect hyperspectral images of the prepared kiwifruit samples. Figure 2 As shown, the hyperspectral imager includes a hyperspectral camera 1, a lens 2, a displacement platform 3, a light source and a scanning controller 4. The hyperspectral imager is equipped with an image acquisition software FigSpec. After system processing, there are 300 available spectral bands between 400-1000nm. The software can automatically perform black and white correction to obtain the corrected spectral reflectance data.

[0060] S3. Segmentation of regions of interest and extraction of spectral data

[0061] In this embodiment, the selection of region of interest (ROI) is necessary in HSI detection and directly affects the performance of the model. ROI is composed of pixels with similar characteristics and is used to extract spectral features. A sample ROI is extracted from the acquired hyperspectral image, and the average spectral information of the region of interest is used as the original spectral information, which is subjected to spectral preprocessing to remove the noise bands at the beginning and end, such as Figure 4 This is the average spectrum of kiwi fruit samples.

[0062] Furthermore, different situations of extracting sample ROI from the acquired hyperspectral image include:

[0063] like Figure 3 As shown in (a), for uninoculated fresh kiwifruit, the part without bumps and any damage was selected as ROI;

[0064] like Figure 3As shown in (b), for naturally infected kiwifruit, the area around the infected point is selected as ROI;

[0065] like Figure 3 As shown in (c), for kiwifruit 7 days after inoculation, the lesion area near the damaged inoculation point is selected as the ROI, where the average reflectance spectrum of the ROI area represents the original spectral characteristics of the sample.

[0066] S4. Spectral data preprocessing

[0067] In this embodiment, during the collection of spectral data, there are some environmental factors such as the instrument itself, sample changes, background noise, optical path scattering, baseline drift, etc., which can make the model unstable, so the spectral data needs to be preprocessed. Different preprocessing methods will affect the modeling, and the use of appropriate spectral preprocessing methods will significantly improve the modeling effect.

[0068] Furthermore, this embodiment uses four preprocessing methods, including Standard Normal Variate (SNV), Multiplicative Scatter Correction (MSC), SG convolution smoothing (Savitzky-Golay) and first-order derivative (First Derivative), to preprocess the original spectrum to reduce the noise caused by various factors. Among them, SNV spectrum preprocessing is mainly used to eliminate the influence of solid particle size on the surface of the object and the scattering of the surface of the object on the diffuse reflectance spectrum; the function of the MSC preprocessing method is similar to that of the SNV spectrum preprocessing, and is mostly used for the diffuse reflectance spectrum of solids, mainly eliminating the spectral scattering effect caused by the uneven distribution of particles on the surface of the object; SG smoothing denoising preprocessing is mainly used to eliminate spectral noise; first-order derivative spectrum preprocessing can effectively eliminate the baseline influence of the collected spectrum and eliminate background interference, and at the same time can effectively distinguish the overlapping peaks of many spectra, thereby improving the specific resolution and sensitivity of the overall model. Spectral preprocessing is performed in The Unscrambler X 10.4. The original (RAW) and the spectra preprocessed by the four methods of SNV, MSC, SG and FD are shown in Figure 1. Figure 5 shown.

[0069] S5. Screening characteristic wavelengths

[0070] In this embodiment, there are many irrelevant collinear variable information in the original spectral data. These redundant variables not only affect the prediction performance of the model, but also greatly increase the calculation time and workload. Through the method of characteristic wavelength screening, the extraction of effective wavelength information can be achieved.

[0071] Furthermore, the present invention adopts genetic algorithm (GA) to extract variables, reduce the negative impact of irrelevant information on prediction performance, and improve the classification accuracy and modeling efficiency of the model. Characteristic wavelength screening can be carried out in MATLAB R2023b. Among them, genetic algorithm is a method for searching for the optimal solution by simulating the natural evolution process. The algorithm uses computer simulation calculations in a mathematical way to convert the problem-solving process into processes such as crossover and mutation of chromosome genes in biological evolution. When solving more complex combinatorial optimization problems, it is usually possible to obtain better optimization results faster than some conventional optimization algorithms.

[0072] S6. Construction and evaluation of disease identification model

[0073] In this embodiment, the specific process of disease identification model construction and evaluation is as follows:

[0074] S61. Use KS algorithm to divide the collected experimental samples into training set: test set = 3:1 (taking this as an example, not limited to this), and obtain 900 training sets and 300 test sets.

[0075] S62, obtain the recognition model by training the selected classifier based on the training set

[0076] In this embodiment, four classification models can be selected as verification, namely, Random Forest (RF), Support Vector Machine (SVM), Back Propagation Neural Network (BPNN) and Extreme Learning Machine (ELM). When selecting a classifier, the diversity and mutual independence of the learner should be fully considered to ensure the effective complementarity of information between the models when constructing an integrated learning model.

[0077] Furthermore, the RF algorithm has high learning efficiency and strong generalization effect, and is suitable for high-dimensional data sets with more features. The number of random trees is set to 100. SVM is a machine learning method based on statistical theory and the principle of structural risk minimization. This method has many advantages in solving sample classification problems, linear, nonlinear and high-dimensional pattern recognition problems. BPNN has the ability to classify arbitrarily complex patterns and excellent multidimensional function mapping capabilities, solving the exclusive OR (XOR) and some other problems that simple perceptrons cannot solve. Structurally, BPNN has an input layer, a hidden layer and an output layer; in essence, the BPNN algorithm uses the square of the network error as the objective function and uses the gradient descent method to calculate the minimum value of the objective function. ELM is a new type of fast learning algorithm. For a single hidden layer neural network, ELM can randomly initialize the input weights and biases and obtain the corresponding output weights. Its biggest feature is that for traditional neural networks, especially single hidden layer feedforward neural networks (SLFNs), it is faster than traditional learning algorithms while ensuring learning accuracy.

[0078] S63. The disease recognition model uses ten-fold cross validation during the training process, and the model is evaluated by the classification accuracy and confusion matrix of the test set. The calculation process of the classification accuracy is as follows:

[0079]

[0080] In the formula, TP and TN represent the positive and negative samples predicted correctly by the model, respectively, FP represents the negative samples predicted by the model as positive, and FN represents the positive samples predicted by the model as negative.

[0081] In summary, the accuracy of spectral data screened by GA characteristic bands through different preprocessing methods and modeling methods is as follows: Figure 6 As shown in the figure, the accuracy of the model is basically the same whether it is screened by GA characteristic bands or not. The accuracy of the model preprocessed by MSC is the highest; except for SVM, the accuracy of the other models is slightly lower, while the accuracy of the other models is above 80%. By comparison, it can be seen that after the original spectrum is preprocessed by MSC, the overall discrimination accuracy of the other models can reach 100.00% except for the RF model.

[0082] Embodiment 2: Embodiment 1 above provides a kiwifruit disease identification method based on hyperspectrum, and correspondingly, this embodiment provides a kiwifruit disease identification device based on hyperspectrum. The device provided in this embodiment can implement the kiwifruit disease identification method based on hyperspectrum of Embodiment 1, and the device can be implemented by software, hardware, or a combination of software and hardware. For the convenience of description, this embodiment is described in terms of functions divided into various units and described separately. Of course, the functions of each unit can be implemented in the same or more software and / or hardware during implementation. For example, the device may include integrated or separate functional modules or functional units to execute the corresponding steps in each method of Embodiment 1. Since the device of this embodiment is basically similar to the method embodiment, the description process of this embodiment is relatively simple, and the relevant parts can refer to the partial description of Embodiment 1. The embodiment of the kiwifruit disease identification device based on hyperspectrum provided by the present invention is only schematic.

[0083] Specifically, the present invention also provides a kiwifruit disease identification device based on hyperspectral, the device comprising:

[0084] A hyperspectral data acquisition unit is configured to acquire and correct a hyperspectral image of a kiwifruit sample to be tested;

[0085] A data segmentation and extraction unit is configured to segment the region of interest and extract spectral data from the corrected hyperspectral image;

[0086] A preprocessing unit, configured to preprocess the extracted spectral data;

[0087] A wavelength screening unit is configured to screen characteristic wavelengths from the preprocessed spectral data through a genetic algorithm;

[0088] The disease identification unit is configured to complete kiwifruit disease identification by using the selected characteristic wavelengths through the constructed disease identification model.

[0089] Embodiment 3: This embodiment provides an electronic device corresponding to the hyperspectral-based kiwi disease identification method provided in Embodiment 1. The electronic device may be an electronic device for a client, such as a mobile phone, a laptop computer, a tablet computer, a desktop computer, etc., to execute the method of Embodiment 1.

[0090] like Figure 7As shown, the electronic device includes a processor, a memory, a communication interface and a bus, and the processor, the memory and the communication interface are connected through the bus to complete mutual communication. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Component (EISA) bus, etc. The memory stores a computer program that can be run on the processor, and the processor executes the method of embodiment 1 when running the computer program. The implementation principle and technical effect are similar to those of embodiment 1, and will not be repeated here. Those skilled in the art can understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present application, and does not constitute a limitation on the computing device to which the scheme of the present application is applied. The specific computing device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0091] In a preferred embodiment, the logic instructions in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), optical disk and other media that can store program codes.

[0092] In a preferred embodiment, the processor may be a central processing unit (CPU), a digital signal processor (DSP) or other general-purpose processors of various types, which are not limited here.

[0093] Embodiment 4: This embodiment provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include computer instructions. When the computer instructions are executed by a computer, the computer executes the method provided in the above-mentioned embodiment 1.

[0094] Embodiment 5: This embodiment provides a computer program product. The computer program product may include a computer program stored on a computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the method provided in the above-mentioned embodiment 1. Its implementation principle and technical effects are similar to those of embodiment 1 and will not be repeated here.

[0095] In a preferred embodiment, the computer-readable storage medium may be a tangible device that holds and stores instructions used by the instruction execution device, such as but not limited to an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof. The computer-readable storage medium stores computer program instructions that cause the computer to execute the method provided in the first embodiment.

[0096] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (apparatus), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0097] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0098] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0099] Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In the description of this specification, the description of reference terms "a preferred embodiment", "further", "specifically", "in the present embodiment", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of this specification. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples without contradiction.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A kiwifruit disease identification method based on hyperspectral, characterized in that: The method includes: Acquire and calibrate the hyperspectral images of the kiwifruit samples to be tested; Segment the region of interest and extract spectral data from the corrected hyperspectral image; Preprocessing the extracted spectral data; The preprocessed spectral data is used to select characteristic wavelengths through a genetic algorithm; The selected characteristic wavelengths are used to complete kiwifruit disease identification through the constructed disease identification model.

2. The kiwifruit disease identification method based on hyperspectral according to claim 1, characterized in that: The FIGSPEC FS10 series hyperspectral imager and the matching image acquisition software FigSpec were used for the acquisition and correction of hyperspectral images.

3. The kiwifruit disease identification method based on hyperspectral according to claim 1, characterized in that: The corrected hyperspectral image is segmented into regions of interest and spectral data is extracted. The specific process is as follows: a sample region of interest ROI is extracted from the acquired hyperspectral image, and the average spectral information of the region of interest is used as the original spectral information, which is subjected to spectral preprocessing to remove the noise bands at both ends.

4. The kiwifruit disease identification method based on hyperspectral according to claim 1, characterized in that: The extracted spectral data were preprocessed using standard normal variable transformation, multivariate scattering correction, SG convolution smoothing and first-order derivative to reduce the noise caused by various factors.

5. The kiwifruit disease identification method based on hyperspectral according to claim 1, characterized in that: The construction of disease identification model includes: preparing kiwi fruit samples; The prepared kiwifruit samples were subjected to hyperspectral image acquisition and correction, region of interest segmentation and spectral data extraction, spectral data preprocessing and characteristic wavelength screening to obtain experimental samples; The KS algorithm is used to divide the experimental samples into training sets and test sets in proportion; The training set is used to train the disease recognition model based on the selected classifier, and the disease recognition model is verified during the training process through the test set.

6. The kiwifruit disease identification method based on hyperspectral according to claim 5, characterized in that: The classifiers used are random forest, support vector machine, back propagation neural network or extreme learning machine.

7. The kiwifruit disease identification method based on hyperspectral according to claim 5, characterized in that: Prepare kiwi fruit samples, including: After cleaning and sterilization, the kiwifruit samples were kept at room temperature of 20°C and relative humidity of RH = 45% for a set time. The selected kiwifruit fruits were inoculated with B. dothidea strains, and a control group without inoculation was set up at the same time; After inoculation, all kiwifruits were moisturized and stored with the inoculated surface facing upward at room temperature of 25°C. The day of inoculation was recorded as day 0, and hyperspectral images were collected when the set number of days after inoculation was met.

8. A kiwifruit disease identification device based on hyperspectral, characterized in that: The device includes: A hyperspectral data acquisition unit is configured to acquire and correct a hyperspectral image of a kiwifruit sample to be tested; A data segmentation and extraction unit is configured to segment the region of interest and extract spectral data from the corrected hyperspectral image; A preprocessing unit, configured to preprocess the extracted spectral data; A wavelength screening unit is configured to screen characteristic wavelengths from the preprocessed spectral data through a genetic algorithm; The disease identification unit is configured to complete kiwifruit disease identification by using the selected characteristic wavelengths through the constructed disease identification model.

9. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor so that the processor can execute the method according to any one of claims 1-7.

10. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include computer instructions for causing a computer to execute the method according to any one of claims 1-7.

Images