Crop growth vigor intelligent sensing algorithm and system, computer equipment and medium

Abstract

The invention discloses an intelligent sensing algorithm and system for crop growth vigor, computer equipment and a medium, and the algorithm comprises the following steps: 1, carrying out the image collection of a farmland, and collecting the spectral data of different wavebands in the farmland; 2, preprocessing the collected images and spectral data, and extracting multi-dimensional feature information of the crops; 3, fusing the extracted multi-dimensional feature information of the crops to obtain comprehensive and accurate crop growth feature information; 4, performing data analysis on the crop growth feature information through deep learning, and constructing an intelligent decision model; through the implementation of the invention, the characteristics and beneficial effects of multi-information fusion, high-precision perception, cost reduction, quality improvement, wide application and the like are realized through the crop growth vigor intelligent perception algorithm, and the method plays an important role in improvement of agricultural production and agricultural product quality.

Description

Technical Field

[0001] The present invention relates to the field of agricultural detection technology, and in particular to an intelligent perception algorithm, system, computer equipment and medium for crop growth. Background Art

[0002] Agriculture is the foundation and important pillar of the national economy, and the growth of crops has an important impact on the benefits of agricultural production and farmers' income. The growth potential of crops refers to the vigor of plants during their growth and development. It is specifically manifested in the growth length, thickness and leaf size of new shoots. The larger, stronger and faster the growth, the stronger the growth potential; on the contrary, if the growth is smaller and weaker, the growth potential is weaker. Growth potential is a direct reflection of the vitality of crops, but it does not have a clear definition, nor does it have a standardized calculation method; leaves, as the place where plants carry out photosynthesis and respiration, are very sensitive to light and the water supply of plants. Therefore, leaf area can be used as an important indicator to measure the growth potential of crops. The larger the leaf area, the more luxuriant the growth and development of the plant, and the better the growth potential, so that more dry matter can be accumulated.

[0003] Traditional crop growth assessment is mainly carried out through manual observation and empirical judgment. This method often has problems such as strong subjectivity, low efficiency and easy misjudgment. The manual method of determining information such as crop plant height requires a lot of manpower time and is not suitable for large-scale monitoring. In addition, some current methods of monitoring crop growth using remote sensing images alone are not comprehensive enough and cannot fully and accurately analyze and calculate the growth conditions of crops in spatial distribution.

[0004] In summary, an intelligent crop growth perception algorithm, system, computer equipment and medium are needed to solve the deficiencies in the prior art. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides an intelligent perception algorithm, system, computer equipment and medium for crop growth, aiming to solve the problems that the above-mentioned traditional methods are highly subjective, inefficient and prone to misjudgment, and are not suitable for large-scale monitoring. In addition, the method of using remote sensing images alone to monitor crop growth is not comprehensive enough and cannot fully and accurately analyze and measure the growth conditions of crops in spatial distribution.

[0006] To achieve the above object, the present invention provides the following technical solution: The present invention relates to an intelligent crop growth perception algorithm, comprising the following steps: Step 1: Collect images of farmland and collect spectral data of different bands in the farmland; Step 2: Preprocess the collected images and spectral data to extract multi-dimensional feature information of crops; Step 3: Fusing the extracted multi-dimensional characteristic information of crops to obtain comprehensive and accurate crop growth characteristic information; Step 4: Analyze crop growth characteristics through deep learning, build an intelligent decision-making model, and iteratively optimize the parameters of the intelligent decision-making model to achieve more accurate prediction and judgment of crop growth; Step 5: According to the crop growth characteristic information and intelligent decision-making model, the newly collected image and spectral data are input into the intelligent decision-making model to obtain the prediction results, and the growth status and health status of the crops are accurately evaluated and monitored. The image and spectral multi-information fusion is used through the crop growth intelligent perception algorithm, which plays the role of multi-information fusion, high-precision perception, cost reduction, quality improvement, and wide application. It can comprehensively monitor the growth of crops, thereby obtaining more comprehensive and accurate crop growth information, which plays an important role in improving agricultural production and the quality of agricultural products.

[0007] Furthermore, in the above-mentioned intelligent perception algorithm for crop growth, in step 2, the collected images and spectral data are preprocessed to extract multidimensional feature information of crops, the image preprocessing includes radiation calibration, atmospheric correction, denoising, enhancement, image correction and image segmentation to extract various feature information of crops, the spectral data preprocessing includes removing stray light, spectral normalization and band selection to extract spectral information of crops, and the spectral information of crops includes NDVI and DVI; Among them, NDVI is the normalized difference vegetation index and DVI is the difference vegetation index.

[0008] Furthermore, in the above-mentioned intelligent perception algorithm for crop growth, in step 3, the extracted multidimensional feature information of crops is integrated to obtain comprehensive and accurate crop growth feature information, wherein the multidimensional feature information includes vegetation coverage, color tone, morphology, growth status, chlorophyll content and physiological indicators.

[0009] Furthermore, in step 3, the extracted multi-dimensional feature information of the crops is fused, and visual features such as shape and texture are extracted from the image using image processing and analysis techniques. At the same time, spectral features are extracted from the spectral data using spectral analysis methods.

[0010] Furthermore, in step 3, the multi-dimensional feature information is fused by using any of the methods of weighted average, principal component analysis, and decision-level fusion to combine and fuse different features to obtain comprehensive crop growth feature information.

[0011] Furthermore, in the above-mentioned crop growth intelligent perception algorithm, in step 4, data analysis is performed on crop growth characteristic information through deep learning to construct an intelligent decision-making model, and the parameters of the intelligent decision-making model are iteratively optimized to achieve more accurate prediction and judgment of crop growth. The data analysis includes the following steps: Step 4.1: Organize the image and spectral data, and label the corresponding crop growth characteristic information according to the fusion features determined in step 3, and divide the data into training set, validation set and test set; Step 4.2: Build a convolutional neural network, design the number of layers, number of neurons and activation function of the network, and select a loss function to measure the performance of the intelligent decision-making model; Step 4.3: Input the preprocessed data set into the convolutional neural network for training, and optimize the intelligent decision-making model during the training process; Step 4.4: Use the validation set to evaluate the model during the training process, optimize the model based on the evaluation results, and output crop growth characteristic information and the iteratively optimized intelligent decision-making model.

[0012] Furthermore, in the above-mentioned intelligent perception algorithm for crop growth, in step 4.4, the crop growth characteristic information and the iteratively optimized intelligent decision-making model are output, and the normalized difference vegetation index is used as the evaluation standard for the intelligent decision-making model to evaluate the crop growth, and the abundance, coverage and crop growth of green vegetation are quantitatively described; The calculation formula of normalized difference vegetation index is: , The calculation formula of difference vegetation index is: , Among them, Rred is the reflectivity in the infrared band, and Rnir is the reflectivity in the near-infrared band.

[0013] Another object of the present invention is to provide an intelligent crop growth perception system, comprising: An acquisition unit, used to obtain high-resolution crop images, spectral images, and spectral data of different bands in the farmland; Image preprocessing unit, used for performing radiometric calibration, atmospheric correction, denoising, enhancement, image correction and image segmentation operations on crop images; A spectral data preprocessing unit is used to remove stray light, perform spectral normalization and band selection on crop spectral images; A feature extraction unit, used to extract multi-dimensional feature information from crop images and spectral data; A feature fusion unit is used to fuse the extracted crop feature information to obtain comprehensive and accurate crop growth feature information; A data analysis unit, used for performing data analysis on crop growth characteristic information; The evaluation and prediction unit is used to accurately evaluate and monitor the growth status and health status of crops.

[0014] Another object of the present invention is to provide a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor performs the following steps: Capture images of farmland, and at the same time, collect spectral data of different bands in the farmland; pre-process the collected images and spectral data to extract multi-dimensional feature information of crops; fuse the extracted multi-dimensional feature information of crops to obtain comprehensive and accurate crop growth feature information; perform data analysis on crop growth feature information through deep learning, build an intelligent decision-making model, and iteratively optimize the parameters of the intelligent decision-making model to achieve more accurate prediction and judgment of crop growth; based on the crop growth feature information and the intelligent decision-making model, input the newly collected images and spectral data into the intelligent decision-making model to obtain prediction results, and accurately evaluate and monitor the growth status and health status of crops.

[0015] Another object of the present invention is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program is executed by a processor to perform the following steps: Capture images of farmland, and at the same time, collect spectral data of different bands in the farmland; pre-process the collected images and spectral data to extract multi-dimensional feature information of crops; fuse the extracted multi-dimensional feature information of crops to obtain comprehensive and accurate crop growth feature information; perform data analysis on crop growth feature information through deep learning, build an intelligent decision-making model, and iteratively optimize the parameters of the intelligent decision-making model to achieve more accurate prediction and judgment of crop growth; based on the crop growth feature information and the intelligent decision-making model, input the newly collected images and spectral data into the intelligent decision-making model to obtain prediction results, and accurately evaluate and monitor the growth status and health status of crops.

[0016] Substantial effects of the present invention: 1. In the present invention, the intelligent perception algorithm of crop growth based on image and spectral multi-information fusion is set up, which has the characteristics and beneficial effects of multi-information fusion, high-precision perception, cost reduction, quality improvement, wide application, etc., and plays an important role in improving agricultural production and the quality of agricultural products.

[0017] 2. In the present invention, a method based on image and spectral multi-information fusion is adopted through the crop growth intelligent perception algorithm to combine image and spectral data, thereby obtaining more comprehensive and accurate crop growth information, so as to achieve multi-information fusion; and, it is possible to perceive the growth status of crops with high precision, and provide scientific basis and decision-making support for agricultural production through analysis and prediction of crop growth, thus playing a role of high-precision perception; 3. In the present invention, by optimizing the traditional agricultural production methods and adopting specific intelligent means, the cost and manpower input of agricultural production are reduced, and the production efficiency and benefits are improved; at the same time, the quality and yield of agricultural products are improved, the comprehensive monitoring and management of the crop growth environment are realized, and the safety and quality of agricultural products are guaranteed, which has made outstanding progress, has strong applicability and flexibility, and has certain use value and promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0019] Figure 1 It is a schematic diagram of the process of the present invention.

[0020] Figure 2 It is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0021] In order to facilitate the understanding of the present invention, the present invention is described in more detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that when an element is described as "fixed to" another element, it can be directly on the other element, or there can be one or more centered elements therebetween. When an element is described as "connected to" another element, it can be directly connected to the other element, or there can be one or more centered elements therebetween. The terms "vertical", "horizontal", "left", "right" and similar expressions used in this specification are for illustrative purposes only.

[0022] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items. Embodiment 1:

[0023] Reference Figure 1 , Figure 2 As shown, the present disclosure relates to an intelligent crop growth perception algorithm, comprising the following steps: Step 1: Use drones and satellites to collect images of farmland to obtain high-resolution crop images. At the same time, collect spectral data of different bands in the farmland to obtain spectral images of crops. Step 2: Preprocess the collected images and spectral data to extract multi-dimensional feature information of crops; These preprocessing operations can fully mine the information in crop images and spectral data, eliminate interference factors, and improve the accuracy and stability of feature extraction. Preprocessing of original remote sensing images includes: establishing a quantitative relationship between digital values ​​and the radiation brightness values ​​in the corresponding field of view through radiometric calibration to eliminate the errors generated by the remote sensing image acquisition sensor itself; restoring the calibration values ​​to the real surface information through atmospheric correction, and restoring the spectral information of the ground objects with high fidelity; removing noise from the image through denoising to reduce errors in subsequent processing; enhancing the image to improve the brightness, contrast and other features of the image to make it clearer and more specific; performing geometric correction and radiometric correction on the image to make it consistent with the earth's surface and accurately express surface features, such as land use type, vegetation coverage, etc.; segmenting the image into a series of regions through image segmentation to extract various characteristic information of crops, such as vegetation index, leaf area index, vegetation coverage, etc. For the collected crop spectral data, the following preprocessing operations are performed: by removing stray light in the spectral data to ensure the accuracy of subsequent processing; By normalizing the spectral data, the differences in spectral reflectance between different times and locations are eliminated, making different spectral data comparable; through band selection, appropriate bands are selected for processing to extract spectral information of crops: the near-infrared band between 0.75 microns and 1.4 microns and the red light band between 0.6 microns and 0.75 microns are selected as vegetation index-related bands, which are very effective in revealing vegetation coverage and growth conditions; the short-wave infrared band between 1.4 microns and 3 microns is selected as the soil-related band, which can be used to identify soil types and obtain information such as soil texture; the blue light band between 0.45 microns and 0.5 microns is selected, which can be used for water body detection and water quality assessment; the red edge band between 0.7 microns and 0.75 microns is selected as the growth status monitoring band, and these bands have good sensitivity for monitoring the chlorophyll content and physiological status of plant leaves.

[0024] Step 3: Fusing the extracted multi-dimensional characteristic information of crops to obtain comprehensive and accurate crop growth characteristic information; Through computer vision technology, multi-dimensional feature information is extracted from crop images and spectral data, and the extracted crop feature information is integrated to obtain comprehensive and accurate crop growth feature information, and weighted average and principal component analysis are specifically used to comprehensively utilize various feature information; Computer vision technology is used to extract multidimensional feature information from crop images and spectral data, and then fuse them to obtain comprehensive and accurate crop growth feature information. Feature extraction refers to extracting representative features based on crop images and spectral data through a series of algorithms and methods. These features may include but are not limited to vegetation coverage, hue, morphology, growth status, chlorophyll content, physiological indicators, etc. In the feature extraction process, image processing and analysis techniques such as image segmentation, edge detection, texture analysis, etc. are used to extract visual features such as shape and texture from the image; at the same time, spectral analysis methods such as vegetation index calculation and spectral curve fitting can be used to extract spectral features from spectral data. Feature fusion is to integrate features extracted from different sources to obtain more comprehensive and accurate feature information. In the feature fusion process, weighted average, principal component analysis, decision-level fusion and other methods are used to combine and fuse different features to obtain comprehensive crop growth feature information. Such feature fusion can make full use of the advantages of different features and improve the stability and discrimination ability of features.

[0025] Step 4: Analyze crop growth characteristics through deep learning, build an intelligent decision-making model, and iteratively optimize the parameters of the intelligent decision-making model to achieve more accurate prediction and judgment of crop growth; Data analysis and model building: Through deep learning, data analysis is performed on crop growth characteristic information, an intelligent decision-making model is constructed, and the parameters of the intelligent decision-making model are iteratively optimized to achieve more accurate prediction and judgment of crop growth; using convolutional neural network deep learning training, data analysis is performed on crop growth characteristic information, model building is performed, an intelligent decision-making model is constructed, and the model parameters are iteratively optimized to achieve more accurate prediction and judgment of crop growth; Establish a convolutional neural network, including input layer, convolution layer, activation layer, pooling layer, batch normalization layer, dropout layer, fully connected layer, and output layer. Each layer considers multiple influencing factors and each layer is composed of multiple neurons. The data set is randomly divided into 70% training set, 20% validation set and 10% test set, and the mean and standard deviation of the sample parameters are calculated: , represents the sample mean of the kth variable, represents the value of the kth variable in the ith observation, represents the sample size, that is, the number of observations, represents the sample standard deviation of the kth variable.

[0026] Change the mean and standard deviation according to the following rules: , represents the standardized value of the kth variable in the ith observation, represents the original value of the kth variable in the ith observation, represents the sample mean of the kth variable, represents the sample standard deviation of the kth variable.

[0027] The influence parameters are normalized to the range [0, 1] by translation transformation: , represents the normalized value of the kth variable in the ith observation, represents the original standardized value of the kth variable in the ith observation, represents the minimum value among all the standardized values ​​of the k-th variable, Represents the maximum value among all the standardized values ​​of the k-th variable.

[0028] Add random perturbations to the training set to enhance the model’s ability to resist interference; The convolutional neural network is trained using the training set. In each round of training, the network parameters are updated using the back propagation algorithm until the preset number of training times is 500 or the correlation coefficient of the test set is greater than 0.95.

[0029] Use the validation set to evaluate the performance of the model, and use regression indicators, mean square error (MSE), and R2 score to measure the prediction accuracy of the model. Optimize the model based on the test results, adjust the network structure, increase training data, modify hyperparameters, etc. until satisfactory results are obtained. Finally, use the test set to finally evaluate the performance of the model and promote it; Step 5: Based on the crop growth characteristic information and the intelligent decision-making model, the newly collected image and spectral data are input into the intelligent decision-making model to obtain the prediction results, and the growth status and health status of the crops are accurately evaluated and monitored; Crop growth assessment and prediction: Based on crop growth characteristic information and intelligent decision-making models, new crop images and spectral image data are input into the network to obtain corresponding NDVI prediction results, which reflect the growth status and health of crops and accurately assess and monitor the growth status and health of crops.

[0030] As an implementation mode, in step 2, the collected images and spectral data are preprocessed to extract multi-dimensional feature information of crops. The image preprocessing includes radiation calibration, atmospheric correction, denoising, enhancement, image correction and image segmentation to extract various feature information of crops. The spectral data preprocessing includes removing stray light, spectral normalization and band selection to extract spectral information of crops. The spectral information of crops includes NDVI and DVI. Among them, NDVI is the normalized difference vegetation index and DVI is the difference vegetation index.

[0031] As an implementation method, in step 3, the extracted multidimensional feature information of crops is integrated to obtain comprehensive and accurate crop growth feature information, wherein the multidimensional feature information includes vegetation coverage, color tone, morphology, growth status, chlorophyll content and physiological indicators.

[0032] As an implementation method, in step 3, the extracted multi-dimensional feature information of the crops is fused, and visual features such as shape and texture are extracted from the image using image processing and analysis techniques. At the same time, spectral features are extracted from the spectral data using spectral analysis methods.

[0033] As an implementation method, in step 3, the multi-dimensional feature information is fused using any of the methods of weighted averaging, principal component analysis, and decision-level fusion to combine and fuse different features to obtain comprehensive crop growth feature information.

[0034] As an implementation mode, in step 4, data analysis is performed on crop growth characteristic information through deep learning to construct an intelligent decision-making model, and the parameters of the intelligent decision-making model are iteratively optimized to achieve a more accurate prediction and judgment of crop growth. The data analysis includes the following steps: Step 4.1: Organize the image and spectral data, and label the corresponding crop growth characteristic information according to the fusion features determined in step 3, and divide the data into training set, validation set and test set; Step 4.2: Build a convolutional neural network, design the number of network layers, number of neurons and activation function, and select a loss function to measure the performance of the intelligent decision-making model. You can choose mean square error, cross entropy, etc. according to the task type; Step 4.3: Input the preprocessed data set into the convolutional neural network for training, and optimize the intelligent decision-making model during the training process; The preprocessed data set is input into the convolutional neural network for training; during the training process, the model should be optimized, such as using optimization algorithms such as gradient descent to update network parameters to minimize the loss function; in order to prevent overfitting, some regularization methods can be used, such as L1 or L2 regularization, dropout, etc.; Step 4.4: Use the validation set to evaluate the model during the training process, optimize the model based on the evaluation results, and output crop growth characteristic information and the iteratively optimized intelligent decision-making model; For example, feature importance analysis can determine which input features are most important for the model's decision-making by analyzing the weights and activation values ​​of each layer in the model. Back-propagation gradient analysis can analyze the gradient information during the back-propagation process to understand which input features are critical to the model output. The relative size and direction of the gradient can be used to determine the contribution of the feature to the model.

[0035] As an implementation method, in step 4.4, the crop growth characteristic information and the iteratively optimized intelligent decision-making model are output; Use the validation set to evaluate the model during the training process to understand the performance of the model, and perform model optimization based on the evaluation results, such as adjusting the network structure, regularization, and hyperparameter adjustment, to output crop growth characteristic information and the iteratively optimized intelligent decision-making model; The normalized difference vegetation index is used as the evaluation standard for the intelligent decision-making model to evaluate the growth of crops, and quantitatively describe the abundance, coverage and growth of green vegetation; The calculation formula of normalized difference vegetation index is: , The calculation formula of difference vegetation index is: , Among them, Rred is the reflectivity in the infrared band, and Rnir is the reflectivity in the near-infrared band.

[0036] Then, the correlation between the relevant vegetation index and the crop growth is statistically analyzed. The correlation analysis method can be: using SPSS data processing software to perform correlation analysis and regression analysis, find out the correlation between the relevant vegetation index with a larger determination coefficient R2 and the crop growth (leaf area index, chlorophyll content, aboveground biomass, plant nitrogen content, etc.) in the analysis results, and use the vegetation index with the highest correlation (the largest determination coefficient R2 and a significant or extremely significant difference) as the basis for band screening Embodiment 2:

[0037] Another aspect of the present invention further provides an intelligent crop growth sensing system, comprising: An acquisition unit, used to obtain high-resolution crop images, spectral images, and spectral data of different bands in the farmland; Image preprocessing unit, used for performing radiometric calibration, atmospheric correction, denoising, enhancement, image correction and image segmentation operations on crop images; A spectral data preprocessing unit is used to remove stray light, perform spectral normalization and band selection on crop spectral images; A feature extraction unit, used to extract multi-dimensional feature information from crop images and spectral data; A feature fusion unit is used to fuse the extracted crop feature information to obtain comprehensive and accurate crop growth feature information; A data analysis unit, used for performing data analysis on crop growth characteristic information; The evaluation and prediction unit is used to accurately evaluate and monitor the growth status and health status of crops. Embodiment 3:

[0038] On the other hand, the present invention also provides a computer device, including a memory and a processor, and a program stored in the processor and executable on the processor. When executing the program, the processor implements the steps of the above-mentioned intelligent perception algorithm for crop growth. Embodiment 4:

[0039] On the other hand, the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned intelligent perception algorithm for crop growth are implemented.

[0040] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a full hardware embodiment, a full software embodiment or a combination of embodiments in terms of hardware and software. In addition, the present invention may also take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0041] It should be noted that the preferred embodiments of the present invention are given in the specification and drawings of the present invention. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in this specification. These embodiments are not intended to be additional limitations on the content of the present invention. The purpose of providing these embodiments is to make the understanding of the disclosure of the present invention more thorough and comprehensive. In addition, the above-mentioned technical features continue to be combined with each other to form various embodiments not listed above, which are all considered to be within the scope of the present invention; further, for ordinary technicians in this field, they can be improved or transformed according to the above description, and all these improvements and transformations should belong to the scope of protection of the claims attached to the present invention.

Claims

1. An intelligent crop growth perception algorithm, characterized in that: The following steps are involved: Step 1: Collect images of farmland and collect spectral data of different bands in the farmland; Step 2: Preprocess the collected images and spectral data to extract multi-dimensional feature information of crops; Step 3: Fusing the extracted multi-dimensional characteristic information of crops to obtain comprehensive and accurate crop growth characteristic information; Step 4: Analyze crop growth characteristics through deep learning, build an intelligent decision-making model, and iteratively optimize the parameters of the intelligent decision-making model to achieve more accurate prediction and judgment of crop growth; Step 5: Based on the crop growth characteristic information and the intelligent decision-making model, the newly collected image and spectral data are input into the intelligent decision-making model to obtain the prediction results, and the growth status and health status of the crops are accurately evaluated and monitored.

2. According to claim 1, the intelligent perception algorithm for crop growth is characterized in that: In the step 2, the collected images and spectral data are preprocessed to extract multi-dimensional feature information of crops, wherein the image preprocessing includes radiation calibration, atmospheric correction, denoising, enhancement, image correction and image segmentation to extract various feature information of crops, and the spectral data preprocessing includes removing stray light, spectral normalization and band selection to extract spectral information of crops, and the spectral information of crops includes NDVI and DVI; Among them, NDVI is the normalized difference vegetation index and DVI is the difference vegetation index.

3. According to claim 1, the intelligent crop growth perception algorithm is characterized in that: In step 3, the extracted multi-dimensional feature information of crops is integrated to obtain comprehensive and accurate crop growth feature information, wherein the multi-dimensional feature information includes vegetation coverage, color tone, morphology, growth status, chlorophyll content and physiological indicators.

4. According to claim 3, the intelligent crop growth perception algorithm is characterized in that: In step 3, the extracted multi-dimensional feature information of the crops is fused, and shape and texture visual features are extracted from the image using image processing and analysis techniques. At the same time, spectral features are extracted from spectral data using spectral analysis methods.

5. According to claim 4, the intelligent crop growth perception algorithm is characterized in that: In step 3, the multi-dimensional feature information is fused by using any method of weighted average, principal component analysis, and decision-level fusion to combine and fuse different features to obtain comprehensive crop growth feature information.

6. The intelligent crop growth perception algorithm according to claim 1, characterized in that: In step 4, data analysis is performed on crop growth characteristic information through deep learning to construct an intelligent decision-making model, and the parameters of the intelligent decision-making model are iteratively optimized to achieve a more accurate prediction and judgment of crop growth. The data analysis includes the following steps: Step 4.1: Organize the image and spectral data, and label the corresponding crop growth characteristic information according to the fusion features determined in step 3, and divide the data into training set, validation set and test set; Step 4.2: Build a convolutional neural network, design the number of layers, number of neurons and activation function of the network, and select a loss function to measure the performance of the intelligent decision-making model; Step 4.3: Input the preprocessed data set into the convolutional neural network for training, and optimize the intelligent decision-making model during the training process; Step 4.4: Use the validation set to evaluate the model during the training process, optimize the model based on the evaluation results, and output crop growth characteristic information and the iteratively optimized intelligent decision-making model.

7. The intelligent crop growth perception algorithm according to claim 6, characterized in that: In step 4.4, the crop growth characteristic information and the iteratively optimized intelligent decision-making model are output, and the normalized difference vegetation index is used as the evaluation standard for the intelligent decision-making model to evaluate the crop growth, and the abundance, coverage and crop growth of green vegetation are quantitatively described; The calculation formula of normalized difference vegetation index is: ， The calculation formula of difference vegetation index is: ， Among them, Rred is the reflectivity in the infrared band, and Rnir is the reflectivity in the near-infrared band.

8. An intelligent crop growth sensing system, characterized in that: include: An acquisition unit, used to obtain high-resolution crop images, spectral images, and spectral data of different bands in the farmland; Image preprocessing unit, used for performing radiometric calibration, atmospheric correction, denoising, enhancement, image correction and image segmentation operations on crop images; A spectral data preprocessing unit is used to remove stray light, perform spectral normalization and band selection on crop spectral images; A feature extraction unit, used to extract multi-dimensional feature information from crop images and spectral data; A feature fusion unit is used to fuse the extracted crop feature information to obtain comprehensive and accurate crop growth feature information; A data analysis unit, used for performing data analysis on crop growth characteristic information; The evaluation and prediction unit is used to accurately evaluate and monitor the growth status and health status of crops.

9. A computer device, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When executing the computer program, the processor implements the steps of the intelligent perception algorithm for crop growth as described in any one of items 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the intelligent perception algorithm for crop growth as described in any one of claims 1 to 7 are implemented.

Images