Database-based Information Processing Method, Apparatus, and Storage Medium

Abstract

This application relates to the field of information processing, and specifically discloses an information processing method, device and storage medium based on a database. First, it obtains the crop growth images collected by a camera and the crop information collected by the database, then uses deep learning technology to perform feature extraction and correlation analysis on the two, and finally passes through a classifier to determine whether it is necessary to adjust the crop growth environment to ensure the healthy growth of the crops, thereby improving the yield and quality of the crops, reducing resource waste at the same time, and realizing sustainable agricultural development.

Description

Technical Field

[0001] This application relates to the field of information processing, and more specifically, to an information processing method, apparatus, and storage medium based on a database. Background Art

[0002] Crops are plants cultivated by humans to meet various needs such as food, feed, textiles, and energy. As the basis of human life and economic development, there are a wide variety of crops, including grains, vegetables, fruits, oil crops, and fiber crops. They require conditions such as photosynthesis, water, and nutrients during the growth process, and different crops have different environmental requirements. Through agricultural activities such as farming, fertilization, irrigation, and pest control, humans continuously improve and cultivate crops to increase yields and quality to meet human needs for food and other uses.

[0003] Crops are divided into food crops and cash crops, among which food crops are related to people's livelihood issues. Scientific cultivation of food crops directly affects the harvest of food, and thus affects food supply. Considering the importance of different growth cycles for crop growth, such as the germination and rooting stage requiring appropriate sunlight and water, while the fruiting stage requires timely fertilization and pest control. Different growth stages have different requirements, and how to coordinate the relationship between fertilization, irrigation, and the natural environment is crucial for crop growth and harvest.

[0004] Currently, the technology of crop planting and fertilization still relies on the setting and implementation of manual experience. Limited by the level of human experience, there are problems of inaccurate control of fertilization amount and time, resulting in waste or affecting crop growth and harvest.

[0005] Therefore, there is a need for an information processing method, apparatus, and storage medium based on a database. Summary of the Invention

[0006] To solve the above technical problems, this application is proposed. Embodiments of this application provide an information processing method, apparatus, and storage medium based on a database. First, it acquires crop growth images collected by a camera and crop information collected by a database, then uses deep learning technology to perform feature extraction and correlation analysis on the two, and finally uses a classifier to determine whether it is necessary to adjust the crop growth environment to ensure the healthy growth of crops, thereby increasing crop yields and quality, while reducing resource waste and achieving sustainable agricultural development.

[0007] According to one aspect of this application, there is provided an information processing method based on a database, which includes:

[0008] Acquire crop growth images collected by a camera and crop information collected by a database;

[0009] Extract an optimized crop growth correlation feature vector and a global crop information correlation feature vector from the crop growth images collected by the camera and the crop information collected by the database;

[0010] Based on the optimized crop growth correlation feature vector and the global crop information correlation feature vector, determine whether it is necessary to adjust the crop growth environment.

[0011] According to another aspect of the present application, there is provided an information processing device based on a database, which includes:

[0012] A crop data acquisition module for acquiring crop growth images collected by a camera and crop information collected by a database;

[0013] A crop data extraction module for extracting an optimized crop growth correlation feature vector and a global crop information correlation feature vector from the crop growth images collected by the camera and the crop information collected by the database;

[0014] A growth environment adjustment module for determining whether it is necessary to adjust the crop growth environment based on the optimized crop growth correlation feature vector and the global crop information correlation feature vector.

[0015] According to still another aspect of the present application, there is provided a computer-readable medium having computer program instructions stored thereon, and when the computer program instructions are run by a processor, the processor is caused to execute the information processing device based on a database as described above.

[0016] Compared with the prior art, an information processing method, device and storage medium based on a database provided by the present application first acquire crop growth images collected by a camera and crop information collected by a database, then use deep learning technology to perform feature extraction and correlation analysis on the two, and finally use a classifier to determine whether it is necessary to adjust the crop growth environment to ensure the healthy growth of crops, thereby improving the yield and quality of crops, while reducing resource waste and realizing sustainable agricultural development. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other objects, features and advantages of the present application will become more obvious. The drawings are used to provide a further understanding of the embodiments of the present application, and constitute a part of the specification, and are used to explain the present application together with the embodiments of the present application, and do not constitute a limitation to the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0018] Figure 1It is a flowchart of an information processing method based on a database according to an embodiment of the present application.

[0019] Figure 2 It is a flowchart of extracting features from the crop growth images collected by a camera in the information processing method based on a database according to an embodiment of the present application to obtain the optimized crop growth association feature vectors.

[0020] Figure 3 It is a flowchart of performing data preprocessing on the crop growth images collected by the camera in the information processing method based on a database according to an embodiment of the present application to obtain a plurality of crop growth image block embedding vectors.

[0021] Figure 4 It is a flowchart of extracting features from the crop information collected by the database in the information processing method based on a database according to an embodiment of the present application to obtain the global association feature vectors of the crop information.

[0022] Figure 5 It is a block diagram of an information processing device based on a database according to an embodiment of the present application.

[0023] Figure 6 It is a block diagram of an electronic device according to an embodiment of the present application. Detailed implementation manners

[0024] Next, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the exemplary embodiments described herein.

[0025] Exemplary method: Figure 1 It is a flowchart of an information processing method based on a database according to an embodiment of the present application. As Figure 1 shown, the information processing method based on a database according to an embodiment of the present application includes: S110, obtaining crop growth images collected by a camera and crop information collected by a database; S120, extracting optimized crop growth association feature vectors and global association feature vectors of crop information from the crop growth images collected by the camera and the crop information collected by the database; S130, based on the optimized crop growth association feature vectors and the global association feature vectors of the crop information, determining whether it is necessary to adjust the crop growth environment.

[0026] In the above information processing method based on a database, in step S110, a crop growth image collected by a camera and crop information collected by the database are obtained. It should be understood that crops, as plants cultivated by humans, play an important role in meeting various human needs such as food, feed, textiles, and energy. They come in a wide variety of types, including grains, vegetables, fruits, oil crops, and fiber crops, and have different requirements for conditions such as photosynthesis, water, and nutrients. Through agricultural activities such as farming, fertilization, irrigation, and pest control, humans continuously strive to improve and cultivate crops, aiming to increase yields and quality to meet human needs for food and other uses. Among them, crops can be divided into food crops and cash crops, and food crops are directly related to people's livelihood issues. Scientific cultivation of food crops directly affects the harvest of food and thus affects food supply. Different growth stages have different requirements for crop growth. For example, the germination and rooting stage requires appropriate sunlight and water, while the fruiting stage requires timely fertilization and pest control. Therefore, coordinating the relationship between fertilization, irrigation, and the natural environment is crucial for crop growth and harvest. However, currently, crop planting and fertilization techniques still rely on the setting and implementation of manual experience, and there are problems with inaccurate control of fertilization amount and timing. This situation leads to resource waste and affects crop growth and harvest. In the technical solution of this application, by obtaining the crop growth image collected by the camera and the crop information collected by the database, and combining deep learning techniques, it is determined whether the crop growth environment needs to be adjusted, thereby optimizing the crop growth process, increasing yields and quality, reducing resource waste at the same time, and promoting the sustainable development of agriculture.

[0027] Specifically, obtaining the crop growth image collected by the camera and the crop information collected by the database can achieve the monitoring and adjustment of the crop growth environment to increase crop yields and quality. Among them, the image collected by the camera provides real-time visual information, which can reflect important information such as the growth status and pest and disease conditions of the crops. By analyzing these images, problems can be discovered in a timely manner and corresponding measures can be taken. The crop information in the database includes information in many aspects such as planting history, growth conditions, and climate data. These data can help better understand the crop growth environment, requirements, and potential problems. By comprehensively analyzing the images collected by the camera and the information in the database, an optimized crop growth correlation feature vector and a crop information global correlation feature vector can be extracted. These feature vectors can help determine whether the current growth status and environment of the crops meet the optimal growth conditions. Based on the analysis of these feature vectors, the crop growth environment can be adjusted in a timely manner, such as adjusting the irrigation water volume, fertilization amount, or pest control, to maximize the growth and development of the crops, thereby increasing crop yields and quality and promoting the sustainable development of agricultural production.

[0028] In the above information processing method based on a database, in step S120, an optimized crop growth correlation feature vector and a global crop information correlation feature vector are extracted from the crop growth images collected by the camera and the crop information collected by the database. It should be understood that the purpose of extracting the optimized crop growth correlation feature vector and the global crop information correlation feature vector from the crop growth images collected by the camera and the crop information collected by the database is to comprehensively utilize visual information and multi-dimensional data to more comprehensively and accurately evaluate the crop growth status and environment, so as to guide agricultural production practices, provide more accurate decision-making basis for agricultural producers, help them timely adjust planting strategies and optimize the growth environment, thereby increasing crop yield, quality and stress resistance, and achieving the goal of sustainable agricultural development.

[0029] In a specific embodiment of the present application, step S120 includes: performing feature extraction on the crop growth images collected by the camera to obtain the optimized crop growth correlation feature vector; performing feature extraction on the crop information collected by the database to obtain the global crop information correlation feature vector.

[0030] It should be understood that the image contains a large amount of visual information, such as color, shape, texture, etc., which can reflect the growth situation and potential problems of the crops. Through feature extraction, these complex visual information can be converted into feature vectors with numerical representations. During the feature extraction process, various computer vision and image processing techniques can be used, such as edge detection, color histogram, texture feature extraction, etc. These techniques can help extract key features in the image, such as vegetation coverage rate, pest and disease locations, growth density, etc. By extracting and analyzing these features, the growth situation of the crops can be understood more accurately, problems can be discovered in a timely manner and corresponding measures can be taken. The obtained optimized crop growth correlation feature vector can contain various information related to crop growth, such as leaf color distribution, vegetation coverage rate, leaf shape, etc. These feature vectors can be used to describe the growth status of the crops and help agricultural experts conduct real-time monitoring and analysis. By comparing the feature vectors at different time points, the growth process of the crops can also be tracked, the growth rate and health status can be evaluated, and a basis for adjusting the growth environment can be provided.

[0031] Furthermore, through feature extraction, the complex information in the database can be converted into representative feature vectors, which contain the key information required for crop growth. For example, features related to soil nutrients, climate conditions, pest and disease control, etc., as well as features related to crop growth cycle, growth rate, yield prediction, etc., can be extracted. These feature vectors can help agricultural experts better understand the characteristics and requirements of the crop growth environment and provide support for decision-making during the crop growth process.

[0032] Figure 2 A flowchart for extracting features from the crop growth images collected by a camera in the information processing method based on a database according to an embodiment of the present application to obtain the optimized crop growth correlation feature vector. As Figure 2 shown, in a specific embodiment of the present application, extracting features from the crop growth images collected by the camera to obtain the optimized crop growth correlation feature vector includes: S210, performing data preprocessing on the crop growth images collected by the camera to obtain a plurality of crop growth image patch embedding vectors; S220, passing the plurality of crop growth image patch embedding vectors through a crop growth image patch converter model based on a ViT model to obtain a crop growth correlation feature vector; S230, based on the global correlation feature vector of the crop information, performing depth response point-by-point correlation compensation in the divergence domain on the crop growth correlation feature vector to obtain an optimized crop growth correlation feature vector.

[0033] It should be understood that in image processing, common preprocessing steps include denoising, image enhancement, size adjustment, etc. Performing data preprocessing on the crop growth images helps to improve the image quality and reduce interference factors. Among them, the process of dividing the preprocessed crop growth images into multiple image patches helps to perform a finer-grained analysis of the images. Each image patch can be regarded as a local area in the image, containing specific crop growth information. By extracting the feature representations of each image patch, the detailed information in the crop growth images, such as leaf shape, color distribution, and disease locations, can be understood in more detail. These embedding vectors can be used for tasks such as image classification, object detection, and image similarity comparison. By comparing the embedding vectors of different image patches, the similarities and differences between images can be found, which helps to identify crops in different growth states and provide more refined growth monitoring and management.

[0034] Furthermore, the embedding vectors of multiple crop growth image patches are transformed through a crop growth image patch transformer model based on the ViT (Vision Transformer) model, aiming to utilize the advantages of the Transformer architecture to convert the spatial information of the image patches into global correlation feature vectors, thereby better capturing the correlation and global information between the image patches. Among them, the ViT model, as a vision processing model based on Transformer, realizes the conversion from the original image to global features by dividing the image into image patches and inputting the embedding vectors of these image patches into the Transformer network. In the processing of crop growth images, this method can help integrate the local features of the image patches into global features, thus better understanding the content and structure of the entire image. Through the crop growth image patch transformer model, the embedding vector of each image patch can capture the semantic correlation and global information between the image patches after being processed and interacted by the Transformer network. By comprehensively using the ViT model and the crop growth image patch transformer model, the local information of the crop growth image patches can be effectively converted into global correlation feature vectors, providing more in-depth and comprehensive visual analysis and decision support for crop growth monitoring and management. Specifically, perform image chunking processing on the embedding vectors of the multiple crop growth image patches to obtain a sequence of image patches; use the embedding layer of the crop growth image patch transformer model based on the ViT model to perform embedding encoding on each image patch in the sequence of image patches to obtain a sequence of image patch embedding vectors; and input the sequence of image patch embedding vectors into the transformer module of the crop growth image patch transformer model based on the ViT model to obtain the crop growth correlation feature vectors.

[0035] Furthermore, based on the global correlation feature vector of the crop information, perform divergence-domain based depth response pointwise correlation compensation on the crop growth correlation feature vector to obtain an optimized crop growth correlation feature vector, including: calculating the position-wise subtraction of the crop growth correlation feature vector and the global correlation feature vector of the crop information to obtain a difference feature vector; dividing the feature values at each position of the difference feature vector by two and then taking the absolute value and calculating the natural exponential function value to obtain an exponentialized feature vector; dividing the second hyperparameter by the feature values of the exponentialized feature vector to obtain a second weighted coefficient vector; multiplying the second weighted coefficient vector with the global correlation feature vector of the crop information position-wise to obtain a weighted global correlation feature vector of the crop information; multiplying the first predetermined hyperparameter with the crop growth correlation feature vector position-wise and then adding it to the weighted global correlation feature vector of the crop information position-wise to obtain the optimized crop growth correlation feature vector.

[0036] In particular, in the technical solution of the present application, it is considered that the crop growth-related feature vector and the global crop information-related feature vector may contain some interrelated information during the extraction process. For example, the growth state of crops may be affected by factors such as planting methods and growth cycles in crop information, resulting in a certain degree of information crossover and duplication in these two feature vectors. During the feature extraction process, there may be some overlapping features that are simultaneously extracted into the crop growth-related feature vector and the global crop information-related feature vector. For example, some features related to the crop growth environment may be reflected in both the crop growth images and crop information, leading to the existence of duplicate or redundant information. There may be some duplicate information in the design of the crop growth-related feature vector and the global crop information-related feature vector. If the model design does not fully consider the differences between the two, it may lead to the existence of duplicate information. Based on this, in the technical solution of the present application, based on the global crop information-related feature vector, a divergence-domain-based depth response pointwise correlation compensation is performed on the crop growth-related feature vector to eliminate the duplicate or redundant parts in the crop growth-related feature vector compared to the global crop information-related feature vector, thereby improving the generalization ability of the crop growth-related feature vector.

[0037] Specifically, based on the global crop information-related feature vector, performing a divergence-domain-based depth response pointwise correlation compensation on the crop growth-related feature vector to obtain an optimized crop growth-related feature vector includes: based on the global crop information-related feature vector, using the following optimization formula to perform a divergence-domain-based depth response pointwise correlation compensation on the crop growth-related feature vector to obtain the optimized crop growth-related feature vector; ; where represents the eigenvalue at the -th position of the crop growth-related feature vector, represents the eigenvalue at the -th position of the global crop information-related feature vector, and represent the first predetermined hyperparameter and the second predetermined hyperparameter, represents multiplication by position, represents addition by position, represents the eigenvalue at the -th position of the optimized crop growth-related feature vector.

[0038] To eliminate the duplicate or redundant parts in the crop growth-related feature vector and the global crop information-related feature vector, and to improve the generalization ability of the crop growth-related feature vector, in the technical solution of this application, based on the global crop information-related feature vector, a depth response point-by-point correlation compensation in the divergence domain is performed on the crop growth-related feature vector. It simulates the depth of the position scattering response of the crop growth-related feature vector and the global crop information-related feature vector in the divergence-like space by calculating the gradient of the variance between the crop growth-related feature vector and the global crop information-related feature vector at each position, and uses the gradient of the variance to construct a depth correlation compensation factor to perform feature depth correlation compensation at each position of the crop growth-related feature vector. In this way, the duplicate or redundant parts in the crop growth-related feature vector and the global crop information-related feature vector are suppressed to improve the generalization ability of the crop growth-related feature vector.

[0039] Figure 3 This is a flowchart of data preprocessing of the crop growth images collected by the camera to obtain multiple crop growth image patch embedding vectors in the information processing method based on a database according to an embodiment of this application. As Figure 3 shown, in a specific embodiment of this application, data preprocessing of the crop growth images collected by the camera to obtain multiple crop growth image patch embedding vectors includes: S310, performing crop growth image segmentation on the crop growth images collected by the camera to obtain a crop growth image patch sequence; S320, performing linear embedding coding on the crop growth image patch sequence to obtain the multiple crop growth image patch embedding vectors.

[0040] It should be understood that by segmenting the crop growth images, complex image information can be decomposed into smaller and more easily processed parts. Each crop growth image patch represents a local area in the image and contains specific crop growth information, such as leaves, fruits, branches, etc. This segmentation method helps to extract local features of the image, capture detailed information, and at the same time reduce the complexity of the entire image. The obtained crop growth image patch sequence can be regarded as a set of local features, and each image patch has its specific position and content. This serialized representation method helps to retain the spatial structure information in the image and can better perform sequence analysis and pattern recognition. By processing the crop growth image patch sequence, it is possible to monitor and analyze changes, trends, and anomalies in the image sequence.

[0041] Furthermore, by performing linear embedding encoding on the sequence of crop growth image patches, the feature information of each image patch can be converted into a continuous and dense vector representation, making it easier to capture and understand the feature relationships between image patches. This encoding method can help extract the semantic information of the image patches, thus better describing the content and features of the image patches. Among them, the process of linear embedding encoding usually involves mapping each image patch in the sequence of crop growth image patches to a high-dimensional space through a linear transformation, and then applying operations such as activation functions to finally obtain embedding vectors of a fixed dimension. These embedding vectors can convert the visual information of the image patches into a numerical representation.

[0042] Figure 4 It is a flowchart for extracting features from the crop information collected by the database to obtain the global associated feature vector of the crop information in the information processing method based on the database according to the embodiments of the present application. As Figure 4 shown, in a specific embodiment of the present application, extracting features from the crop information collected by the database to obtain the global associated feature vector of the crop information includes: S410, passing the crop information collected by the database through a crop information text understanding model to obtain a plurality of crop information data items; S420, performing two-dimensional arrangement on the plurality of crop information data items to obtain a crop information associated feature matrix; S430, passing the crop information associated feature matrix through a crop information associated convolutional neural network as a feature encoder to obtain the global associated feature vector of the crop information.

[0043] It should be understood that the application of the crop information text understanding model can help the system automatically parse and understand a large amount of crop information text, and extract useful information from it, such as key data items such as crop species, growth stages, pest and disease situations, and climate elements. This automated processing method can greatly improve the efficiency and accuracy of data processing. Through the crop information text understanding model, semantic understanding and information extraction of text data can be achieved. The model will identify keywords, phrases, and sentence structures in the text, and then map them to predefined data items or entities. In this way, the original text data is converted into a series of structured data items, which is convenient for the system to further process and utilize. Among them, the design of the crop information text understanding model usually involves natural language processing technologies, such as word embedding, named entity recognition, keyword extraction, semantic analysis, etc. These technologies can help the model understand the meaning of the text, identify important information in it, and convert it into a data representation form that is easy to process. By processing the crop information collected by the database through the crop information text understanding model, intelligent management and analysis of crop data can be realized.

[0044] Furthermore, the process of two-dimensionally arranging multiple crop information data items to obtain a crop information correlation feature matrix aims to clearly present the correlation and connection between different crop information in the form of a matrix. In the crop information correlation feature matrix, each row or column usually represents a specific crop information data item, and each element in the matrix represents the degree of association or correlation between the corresponding information items. By two-dimensionally arranging multiple crop information data items, different information items can be organized in a matrix according to a certain order and structure, making the association between information more intuitive and easy to understand. This arrangement helps to present the connection and interaction between different information items, providing a better basis for further data analysis and mining.

[0045] Furthermore, passing the crop information correlation feature matrix through a crop information correlation convolutional neural network as a feature encoder to obtain a crop information global correlation feature vector is to better capture the global correlation features between crop information, thereby improving the effect of data processing and analysis. It should be understood that the crop information correlation feature matrix contains the correlation between different crop information data items. By two-dimensionally arranging these data items to form a matrix, the relationship between different data items can be presented more clearly, so as to extract spatial information from the feature matrix using convolutional operations and realize the transmission and integration of information through the connection structure of the network layers. Among them, the crop information correlation convolutional neural network as a feature encoder can effectively learn and extract the global correlation features between crop information data items. Through convolutional operations and the hierarchical structure of the neural network, the network can automatically learn the feature patterns in the data and encode these feature information into more representative and high-level feature representations to achieve the extraction of global correlation features of crop information, so as to more comprehensively understand and utilize crop information data and provide more accurate and effective support for decision-making and applications in the agricultural field. Specifically, each layer of the crop information correlation convolutional neural network used as the feature encoder performs convolutional processing, mean pooling processing based on the local feature matrix, and non-linear activation processing on the input data respectively during the forward pass of the layer to output the crop information global correlation feature vector by the last layer of the crop information correlation convolutional neural network used as the feature encoder, where the input of the crop information correlation convolutional neural network used as the feature encoder is the crop information correlation feature matrix.

[0046] In the above information processing method based on a database, in step S130, based on the optimized crop growth correlation feature vector and the global crop information correlation feature vector, it is determined whether it is necessary to adjust the crop growth environment. It should be understood that the optimized crop growth correlation feature vector contains various feature information related to crop growth, such as growth rate, leaf color, plant height, etc. These feature vectors can reflect the current growth state and trend of the crop, helping to understand the overall growth situation of the crop. Secondly, the global crop information correlation feature vector provides a more comprehensive data representation of crop information, including various environmental factors and management measures related to crop growth. By analyzing the global correlation feature vector, various influencing factors in the crop growth environment, such as temperature, humidity, light, etc., as well as the situation of management measures such as fertilization and irrigation, can be understood. Combining the information of these two feature vectors, the growth status of the crop and the impact of the growth environment can be comprehensively evaluated, so as to determine whether it is necessary to adjust the crop growth environment. For example, if the optimized crop growth correlation feature vector shows a decrease in growth rate, abnormal leaves, etc., and the global correlation feature vector shows problems such as too high environmental temperature or lack of nutrients in the soil, corresponding adjustment measures may need to be taken, such as adjusting the temperature, increasing the fertilization amount, etc., to improve the crop growth environment and promote healthy growth. By comprehensively using the optimized crop growth correlation feature vector and the global crop information correlation feature vector, the status of the crop growth environment can be judged more accurately, problems can be discovered in time and effective adjustment measures can be taken, so as to improve the crop yield and quality and realize the sustainable development and optimized management of agricultural production.

[0047] In a specific embodiment of the present application, in step S130, determining whether it is necessary to adjust the crop growth environment includes: fusing the optimized crop growth correlation feature vector and the global crop information correlation feature vector to obtain a crop growth environment adjustment judgment feature vector; passing the crop growth environment adjustment judgment feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether it is necessary to adjust the crop growth environment.

[0048] It should be understood that fusing the optimized crop growth-related feature vector and the global crop information-related feature vector to obtain the crop growth environment adjustment judgment feature vector is to comprehensively consider the crop growth situation and growth environment factors, so as to more accurately evaluate whether it is necessary to adjust the crop growth environment. By fusing these two feature vectors, the impacts of the crop growth situation and growth environment factors can be comprehensively considered, and a more comprehensive crop growth environment adjustment judgment feature vector can be obtained. The fused feature vector will integrate the information of the crop's own growth characteristics and the surrounding environment factors, and can more comprehensively reflect the overall situation of crop growth. The application of this comprehensive feature vector can help agricultural producers more accurately judge the status of the crop growth environment, timely discover potential problems and take corresponding adjustment measures to optimize the crop growth environment and improve the yield and quality.

[0049] Furthermore, by fusing the optimized crop growth-related feature vector and the global crop information-related feature vector into a comprehensive feature vector, a comprehensive feature representation of the crop growth situation and growth environment is obtained. This feature vector contains rich information, covering all aspects of crop growth and the impacts of environmental factors. Then, this comprehensive feature vector is input into a classifier for training and classification. The classifier can be various machine learning models, such as support vector machines, decision trees, neural networks, etc. In the training stage, the classifier will learn the relationships between the feature vector and different categories, thus establishing a classification model. Once the classifier is trained, new feature vectors can be input into the classifier to obtain corresponding classification results. Finally, by analyzing the results output by the classifier, it can be judged whether it is necessary to adjust the crop growth environment. By comprehensively using feature extraction, machine learning classification, and decision analysis, classifying the crop growth environment adjustment judgment feature vector through the classifier provides a scientific and intelligent decision-making support method for agricultural production, which helps to achieve precision agriculture management and improve crop production efficiency and quality.

[0050] In summary, in the embodiment of the present application, first, the crop growth images collected by the camera and the crop information collected by the database are obtained, then deep learning technology is used to perform feature extraction and correlation analysis on the two, and finally, through the classifier, it is judged whether it is necessary to adjust the crop growth environment to ensure the healthy growth of the crop, thereby improving the yield and quality of the crop, while reducing resource waste and realizing sustainable agricultural development.

[0051] Exemplary device: Figure 5 It is a block diagram of an information processing device based on a database according to an embodiment of the present application. As Figure 5As shown, the information processing device 100 based on a database according to an embodiment of the present application includes: a crop data acquisition module 110 for acquiring crop growth images collected by a camera and crop information collected by the database; a crop data extraction module 120 for extracting an optimized crop growth correlation feature vector and a crop information global correlation feature vector from the crop growth images collected by the camera and the crop information collected by the database; and a growth environment adjustment module 130 for determining whether it is necessary to adjust the crop growth environment based on the optimized crop growth correlation feature vector and the crop information global correlation feature vector.

[0052] Here, those skilled in the art can understand that the specific operations of each step in the above information processing device based on a database have been described in detail in the description of the Figures 1 to 4 information processing method based on a database above, and therefore, the repeated description thereof will be omitted.

[0053] As described above, the information processing device 100 based on a database according to an embodiment of the present application can be implemented in various terminal devices. In one example, the information processing device 100 based on a database can be integrated into a terminal device as a software module and / or a hardware module. For example, the information processing device 100 based on a database can be a software module in the operating system of the terminal device, or can be an application program developed for the terminal device; of course, the information processing device 100 based on a database can also be one of many hardware modules of the terminal device.

[0054] Alternatively, in another example, the information processing device 100 based on a database and the terminal device can also be separate devices, and the information processing device 100 based on a database can be connected to the terminal device through a wired and / or wireless network and transmit and interact information in accordance with a predefined data format.

[0055] Exemplary electronic device and its storage medium: Next, reference Figure 6 is made to describe the electronic device according to an embodiment of the present application.

[0056] As Figure 6 shown, the electronic device 10 includes an input device 11, an input interface 12, a central processing unit 13, a memory 14, an output interface 15, an output device 16, and a bus 17. Among them, the input interface 12, the central processing unit 13, the memory 14, and the output interface 15 are connected to each other through the bus 17, and the input device 11 and the output device 16 are respectively connected to the bus 17 through the input interface 12 and the output interface 15, and further connected to other components of the electronic device 10.

[0057] Specifically, the input device 11 receives input information from the outside and transmits the input information to the central processing unit 13 through the input interface 12; the central processing unit 13 processes the input information based on computer-executable instructions stored in the memory 14 to generate output information, stores the output information temporarily or permanently in the memory 14, and then transmits the output information to the output device 16 through the output interface 15; the output device 16 outputs the output information to the outside of the electronic device 10 for use by the user.

[0058] In one embodiment, Figure 6 The illustrated electronic device 10 may be implemented as a network device, and the network device may include: a memory configured to store a program; a processor configured to run the program stored in the memory to execute any one of the information processing methods based on a database described in the above embodiments.

[0059] According to an embodiment of the present application, the process described above with reference to the flowchart may be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program tangibly embodied on a machine-readable medium, and the computer program includes program code for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network, and / or installed from a removable storage medium.

[0060] Those of ordinary skill in the art will understand that all or some of the steps in the methods disclosed above, and the functional modules / units in systems and devices, can be implemented as software, firmware, hardware, and their appropriate combinations. In the hardware implementation, the division between the functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component can have multiple functions, or a function or step can be executed by several physical components in cooperation. Some or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include a computer storage medium (or non-transitory medium) and a communication medium (or transitory medium). As is well known to those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassette, tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, as is well known to those of ordinary skill in the art, a communication medium typically contains computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and can include any information delivery medium.

[0061] It can be understood that the above embodiments are merely exemplary embodiments adopted to illustrate the principles of the present application, but the present application is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present application, and these modifications and improvements are also considered within the protection scope of the present application.

Claims

1. A database-based information processing method, characterized in that: include: Acquire crop growth images collected by the camera and crop information collected by the database; Extracting optimized crop growth associated feature vectors and crop information global associated feature vectors from the crop growth images collected by the camera and the crop information collected by the database; Based on the optimized crop growth-related feature vector and the crop information global-related feature vector, determining whether it is necessary to adjust the crop growth environment; The step of extracting optimized crop growth correlation feature vectors and crop information global correlation feature vectors from the crop growth images collected by the camera and the crop information collected by the database includes: Performing feature extraction on the crop growth image captured by the camera to obtain the optimized crop growth associated feature vector; Performing feature extraction on the crop information collected from the database to obtain a global correlation feature vector of the crop information; The step of extracting features from the crop growth image captured by the camera to obtain the optimized crop growth-related feature vector includes: Performing data preprocessing on the crop growth image collected by the camera to obtain a plurality of crop growth image block embedding vectors; Passing the plurality of crop growth image block embedding vectors through a crop growth image block converter model based on a ViT model to obtain a crop growth associated feature vector; Based on the crop information global correlation feature vector, performing divergence-domain-based deep response point-by-point correlation compensation on the crop growth correlation feature vector to obtain an optimized crop growth correlation feature vector; Wherein, based on the crop information global correlation feature vector, the crop growth correlation feature vector is subjected to a depth response point-by-point correlation compensation based on a divergence domain to obtain an optimized crop growth correlation feature vector, including: Calculating the positional subtraction between the crop growth-related feature vector and the crop information global-related feature vector to obtain a phase difference feature vector; Dividing the eigenvalues ​​of each position of the phase difference eigenvector by two and taking the absolute value, and then calculating the natural exponential function value to obtain an exponential eigenvector; Dividing a second hyperparameter by each eigenvalue of the exponentially modified eigenvector to obtain a second weighting coefficient vector; Multiplying the second weighting coefficient vector by the crop information global correlation feature vector by position to obtain a weighted crop information global correlation feature vector; The first predetermined hyperparameter is positionally multiplied with the crop growth associated feature vector and then positionally added to the weighted crop information global associated feature vector to obtain the optimized crop growth associated feature vector.

2. The database-based information processing method according to claim 1, characterized in that: The crop growth image captured by the camera is subjected to data preprocessing to obtain a plurality of crop growth image block embedding vectors, including: dividing the crop growth image captured by the camera into crop growth image blocks to obtain a crop growth image block sequence; The crop growth image block sequence is subjected to crop growth image block sequence linear embedding coding to obtain the plurality of crop growth image block embedding vectors.

3. The method for information processing based on a database according to claim 2, characterized in that: Extracting features of the crop information collected from the database to obtain a global correlation feature vector of the crop information includes: The crop information collected from the database is passed through a crop information text understanding model to obtain a plurality of crop information data items; Arranging the plurality of crop information data items in two dimensions to obtain a crop information associated feature matrix; The crop information association feature matrix is ​​passed through a crop information association convolutional neural network as a feature encoder to obtain the crop information global association feature vector.

4. The database-based information processing method according to claim 3, characterized in that: Based on the optimized crop growth associated feature vector and the crop information global associated feature vector, determining whether it is necessary to adjust the crop growth environment includes: The optimized crop growth associated feature vector and the crop information global associated feature vector are merged to obtain a crop growth environment adjustment judgment feature vector; The crop growth environment adjustment judgment feature vector is passed through a classifier to obtain a classification result, and the classification result is used to determine whether the crop growth environment needs to be adjusted.

5. A database-based information processing device, characterized in that: include: A crop data acquisition module, used to acquire crop growth images collected by a camera and crop information collected by a database; A crop data extraction module, used for extracting optimized crop growth correlation feature vectors and crop information global correlation feature vectors from the crop growth images collected by the camera and the crop information collected by the database; A growth environment adjustment module, used for judging whether it is necessary to adjust the growth environment of crops based on the optimized crop growth associated feature vector and the crop information global associated feature vector; Wherein, the crop data extraction module includes: Performing feature extraction on the crop growth image captured by the camera to obtain the optimized crop growth associated feature vector; Performing feature extraction on the crop information collected from the database to obtain a global correlation feature vector of the crop information; The step of extracting features from the crop growth image captured by the camera to obtain the optimized crop growth-related feature vector includes: Performing data preprocessing on the crop growth image collected by the camera to obtain a plurality of crop growth image block embedding vectors; Passing the plurality of crop growth image block embedding vectors through a crop growth image block converter model based on a ViT model to obtain a crop growth associated feature vector; Based on the crop information global correlation feature vector, performing divergence-domain-based deep response point-by-point correlation compensation on the crop growth correlation feature vector to obtain an optimized crop growth correlation feature vector; Wherein, based on the crop information global correlation feature vector, the crop growth correlation feature vector is subjected to a depth response point-by-point correlation compensation based on a divergence domain to obtain an optimized crop growth correlation feature vector, including: Calculating the positional subtraction between the crop growth-related feature vector and the crop information global-related feature vector to obtain a phase difference feature vector; Dividing the eigenvalues ​​of each position of the phase difference eigenvector by two and taking the absolute value, and then calculating the natural exponential function value to obtain an exponential eigenvector; Dividing a second hyperparameter by each eigenvalue of the exponentially modified eigenvector to obtain a second weighting coefficient vector; Multiplying the second weighting coefficient vector by the crop information global correlation feature vector by position to obtain a weighted crop information global correlation feature vector; The first predetermined hyperparameter is positionally multiplied with the crop growth associated feature vector and then positionally added to the weighted crop information global associated feature vector to obtain the optimized crop growth associated feature vector.

6. A readable storage medium, wherein a plurality of instructions are stored in the readable storage medium; the plurality of instructions are used for a processor to load and execute the database-based information processing method as described in any one of claims 1 to 4.

Images