Intelligent file cabinet state monitoring system and method based on deep learning

Abstract

The invention relates to the field of file cabinet monitoring, and particularly discloses an intelligent file cabinet state monitoring system and method based on deep learning, and the method comprises the steps: firstly obtaining a file cabinet internal state monitoring image collected by a camera installed in a file cabinet and environment parameters collected by an environment integrated sensor, and then employing a deep learning technology to carry out the monitoring of the internal state of the file cabinet; and feature extraction and correlation analysis are carried out on the two files, and finally, whether the file cabinet dehumidification operation needs to be started is judged through a classifier, so that the potential humidity problem is found and processed in time, the files are effectively protected from being affected by moisture and damage, and the file management efficiency and quality are improved.

Description

Technical Field

[0001] The present application relates to the field of file cabinet monitoring, and more specifically, to a deep learning-based intelligent file cabinet status monitoring system and method. Background Art

[0002] A filing cabinet is a piece of furniture specifically used to store files or documents. It is usually made of metal or wood and is designed with multiple drawers or compartments inside to organize and manage files for easy access at any time.

[0003] However, there are some challenges when using filing cabinets in the southern region where humidity is high. Frequent opening of the cabinet door to search for files may cause the cabinet door to be left open for a long time, which can easily cause the files in the cabinet to get damp, leading to mold problems. The traditional solution is to perform dehumidification operations regularly, but this method not only wastes energy, but also cannot be adjusted intelligently according to actual weather conditions, which may lead to unnecessary consumption of energy resources.

[0004] Therefore, a smart file cabinet status monitoring system and method based on deep learning is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a deep learning-based intelligent file cabinet status monitoring system and method, which first obtains the file cabinet internal status monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor, and 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 turn on the file cabinet dehumidification operation, so as to timely discover and deal with potential humidity problems, effectively protect files from moisture and damage, and thus improve the efficiency and quality of file management.

[0006] According to one aspect of the present application, a deep learning-based intelligent file cabinet status monitoring system is provided, which includes:

[0007] A humidity status data acquisition module, used to acquire the internal status monitoring image of the filing cabinet collected by the camera installed inside the filing cabinet and the environmental parameters collected by the environmental integrated sensor;

[0008] A humidity state data extraction module, used to extract a file cabinet internal state monitoring area associated feature vector and a file cabinet environmental parameter feature vector from the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor;

[0009] The file cabinet dehumidification judgment module is used to judge whether it is necessary to start the file cabinet dehumidification operation based on the associated feature vector of the file cabinet internal state monitoring area and the file cabinet environmental parameter feature vector.

[0010] According to another aspect of the present application, a method for monitoring the state of an intelligent file cabinet based on deep learning is provided, which includes:

[0011] Acquire the internal state monitoring image of the filing cabinet collected by the camera installed inside the filing cabinet and the environmental parameters collected by the environmental integrated sensor;

[0012] Extracting a file cabinet internal state monitoring area associated feature vector and a file cabinet environmental parameter feature vector from the file cabinet internal state monitoring image acquired by the camera installed inside the file cabinet and the environmental parameters acquired by the environmental integrated sensor;

[0013] The file cabinet determines whether it is necessary to start the dehumidification operation of the file cabinet based on the associated feature vector of the file cabinet internal state monitoring area and the feature vector of the file cabinet environment parameter.

[0014] Compared with the prior art, the present application provides an intelligent filing cabinet status monitoring system and method based on deep learning, which first obtains the filing cabinet internal status monitoring image collected by the camera installed inside the filing cabinet and the environmental parameters collected by the environmental integrated sensor, and 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 start the filing cabinet dehumidification operation, so as to timely discover and deal with potential humidity problems, effectively protect files from moisture and damage, and thus improve the efficiency and quality of file management. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 4 is a block diagram of a smart file cabinet status monitoring system based on deep learning according to an embodiment of the present application.

[0017] Figure 2 This is a block diagram of a humidity status data extraction module in a deep learning-based intelligent file cabinet status monitoring system according to an embodiment of the present application.

[0018] Figure 3 It is a block diagram of a file cabinet internal state feature extraction unit in a deep learning-based intelligent file cabinet state monitoring system according to an embodiment of the present application.

[0019] Figure 4This is a block diagram of a file cabinet dehumidification judgment module in a deep learning-based intelligent file cabinet status monitoring system according to an embodiment of the present application.

[0020] Figure 5 This is a flowchart of a method for monitoring the status of an intelligent file cabinet based on deep learning according to an embodiment of the present application.

[0021] Figure 6 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0023] Figure 1 : is a block diagram of a smart file cabinet status monitoring system based on deep learning according to an embodiment of the present application. Figure 1 As shown, according to the deep learning-based intelligent file cabinet status monitoring system 100 of the embodiment of the present application, it includes: a humidity status data acquisition module 110, which is used to obtain the file cabinet internal status monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor; a humidity status data extraction module 120, which is used to extract the file cabinet internal status monitoring area associated feature vector and the file cabinet environmental parameter feature vector from the file cabinet internal status monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor; a file cabinet dehumidification judgment module 130, which is used to judge whether it is necessary to start the file cabinet dehumidification operation based on the file cabinet internal status monitoring area associated feature vector and the file cabinet environmental parameter feature vector.

[0024] In the above-mentioned intelligent file cabinet state monitoring system 100 based on deep learning, the humidity state data acquisition module 110 is used to obtain the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor. It should be understood that the file cabinet is a piece of furniture designed to store files and documents, usually made of metal or wood, with multiple drawers or compartments inside, which is convenient for organizing and managing files and convenient for access at any time. However, there are some challenges when using file cabinets in southern regions with high humidity. Frequent opening of the cabinet door to find files may cause the cabinet door to be open for a long time, which will increase the risk of files in the cabinet getting damp, and may cause mold problems. The traditional solution is to perform dehumidification operations regularly, but this method not only wastes energy, but also cannot be adjusted intelligently according to actual weather conditions, which may lead to unnecessary resource consumption. Therefore, in the technical solution of the present application, by obtaining the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor, and combining deep learning technology, it is determined whether the file cabinet dehumidification operation needs to be turned on, effectively protecting the files from moisture and damage, thereby improving the efficiency and quality of file management.

[0025] Specifically, the internal status monitoring images of the filing cabinets collected by the camera provide intuitive visual information. Through these images, the system can observe the layout of the filing cabinets, the storage conditions, and possible humidity problems in real time. The camera can capture subtle details, such as the opening and closing status of each filing cabinet, the utilization of the internal space, etc., which are crucial for evaluating the humidity problems of the filing cabinets and determining whether dehumidification operations are required. The environmental parameters collected by the environmental integrated sensor provide real-time data on the environment around the filing cabinet. These parameters include temperature, humidity, air quality, etc., which directly affect the humidity changes inside the filing cabinet and the safety of the stored items. For example, a high humidity environment can cause the documents and papers to become damp, deformed, or even moldy, while proper humidity control can maintain the integrity and readability of the documents. The monitoring of environmental parameters enables the system to make corresponding adjustments and decisions on a real-time basis to maintain the appropriate environmental conditions inside the filing cabinet. By comprehensively utilizing camera images and environmental sensor data, the intelligent filing cabinet status monitoring system can accurately identify humidity problems and potential risks inside the filing cabinet, provide timely warnings and responses, and improve the management efficiency and document protection capabilities of the filing cabinet, effectively extending the service life of the documents and maintaining their availability.

[0026] In the above-mentioned intelligent file cabinet state monitoring system 100 based on deep learning, the humidity state data extraction module 120 is used to extract the associated feature vector of the file cabinet internal state monitoring area and the file cabinet environmental parameter feature vector from the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor. It should be understood that by comprehensively utilizing the feature vectors extracted from the camera monitoring image and the environmental sensor data, the intelligent file cabinet state monitoring system can accurately identify the humidity problems and potential risks inside the file cabinet, issue early warnings in time, and intelligently adjust the humidity control strategy of the file cabinet according to real-time data, thereby improving the intelligent level of operation, thereby optimizing the space utilization and file storage layout inside the file cabinet, and improving the file management efficiency and file protection capabilities.

[0027] Figure 2 FIG. 4 is a block diagram of a humidity status data extraction module in a smart file cabinet status monitoring system based on deep learning according to an embodiment of the present application. Figure 2 As shown, in a specific embodiment of the present application, the humidity status data extraction module 120 includes: a file cabinet internal status feature extraction unit 121, which is used to perform feature extraction on the file cabinet internal status monitoring image collected by the camera installed inside the file cabinet to obtain the file cabinet internal status monitoring area associated feature vector; a file cabinet environmental parameter feature extraction unit 122, which is used to perform feature extraction on the environmental parameters collected by the environmental integrated sensor to obtain the file cabinet environmental parameter feature vector.

[0028] It should be understood that the surveillance images collected by the camera provide visual information about the real-time status of the interior of the filing cabinet. Through feature extraction, the system can capture and analyze various features of the filing cabinet from the image, such as the opening and closing status of the filing cabinet, the layout and density of the files, possible damage or abnormal conditions, etc. These feature vectors can help the system accurately identify and describe the current status of the filing cabinet, thereby realizing intelligent monitoring and management of the file storage environment.

[0029] Furthermore, environmental integrated sensors are often used to monitor environmental parameters around file cabinets, such as temperature, humidity, and air pressure. Through feature extraction, representative and key features can be extracted from the raw data collected by the sensor, such as environmental change trends, periodic changes, abnormal conditions, etc. These features can help the system understand and describe the environmental conditions of the file cabinet more accurately, so as to promptly discover and solve problems that may affect the quality and storage status of the files. By extracting the feature vectors of environmental parameters, the system can establish a correlation model between environmental conditions and file preservation quality, and then formulate and optimize file storage management strategies suitable for the current environment to ensure that the internal environment of the file cabinet is always in the best state.

[0030] Figure 3 FIG. 1 is a block diagram of a file cabinet internal state feature extraction unit in a smart file cabinet state monitoring system based on deep learning according to an embodiment of the present application. Figure 3 As shown, in a specific embodiment of the present application, the file cabinet internal state feature extraction unit 121 includes: a file cabinet internal humidity state preprocessing subunit 1211, which is used to perform data preprocessing on the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet to obtain a file cabinet internal humidity state monitoring area image block sequence; a file cabinet internal state feature encoding subunit 1212, which is used to perform feature encoding on the file cabinet internal humidity state monitoring area image block sequence to obtain the file cabinet internal state monitoring area associated feature vector.

[0031] It should be understood that the images captured by surveillance cameras usually contain a lot of irrelevant information, such as background, non-interest areas, etc. The first step of data preprocessing is to extract the humidity monitoring areas related to the inside of the filing cabinet through image segmentation or region detection technology. These areas are usually the internal space of the filing cabinet, with special attention paid to the area of ​​documents and other storage items. From the extracted area of ​​interest, the system needs to further process the image data to accurately capture the details of each image block. The processed image is divided into a sequence of image blocks. Each image block represents the local information of a specific area inside the filing cabinet. This segmentation enables the system to perform separate humidity status analysis and monitoring on different parts of the filing cabinet. Each image block can be regarded as an independent data unit, which helps the system to accurately track and record the humidity changes in different areas inside the filing cabinet. The obtained image block sequence of the humidity status monitoring area of ​​the filing cabinet can be used for further humidity analysis and model building. By analyzing the color, texture and other visual features of each image block, the system can infer the humidity status of each area and monitor the humidity change trend in real time. This monitoring method based on image block sequence can help the system to timely detect and deal with potential humidity problems and protect the integrity and long-term preservation quality of documents.

[0032] Furthermore, feature encoding is performed on the image block sequence of the humidity status monitoring area inside the filing cabinet, aiming to convert complex image information into numerical representations that can be used for analysis and decision-making by extracting and encoding key visual features. Each image block in the image block sequence represents local information of different areas inside the filing cabinet, especially information about the humidity status. The first step of feature encoding is to extract relevant visual features from each image block, such as color histogram, texture features, edge information, etc. These features can help the system capture the humidity-related visual performance of each area, such as the degree of humidity, possible signs of mold, etc. The extracted visual features need to be encoded and normalized to generate representative and statistically significant feature vectors. The encoding process can use techniques such as principal component analysis (PCA), local binary pattern (LBP), and histogram of oriented gradients (HOG) to convert high-dimensional feature data into feature vectors with lower dimensions but retaining key information. This encoding method not only helps to reduce data complexity and storage requirements, but also improves the efficiency of subsequent analysis and model building.

[0033] In a specific embodiment of the present application, the filing cabinet internal humidity status preprocessing subunit 1211 includes: performing region segmentation on the filing cabinet internal status monitoring image captured by a camera installed inside the filing cabinet to obtain a plurality of filing cabinet internal humidity status monitoring region images; arranging the plurality of filing cabinet internal humidity status monitoring region images to obtain a sequence of image blocks of the filing cabinet internal humidity status monitoring region.

[0034] It should be understood that the image of the internal state of the filing cabinet captured by the surveillance camera usually contains the field of view of the entire filing cabinet, including the file storage area, the free area, and possible walls or other objects. Through region segmentation, the entire image can be divided into multiple small areas or image blocks, each of which represents a specific part of the inside of the filing cabinet. Region segmentation technology can help identify and extract specific areas related to humidity monitoring inside the filing cabinet. For example, the area where files are stored in the filing cabinet can be automatically identified based on the color, texture or shape features in the image. These areas are often the parts that need humidity monitoring and management the most. The segmented image area can more accurately reflect the changes in the humidity state of each area. Among them, each image area can be regarded as an independent monitoring unit, and the system can analyze the humidity change trend, abnormal state or other important information in each area separately. This refined monitoring method helps to timely discover and deal with humidity problems that may affect the quality of file protection.

[0035] Furthermore, by arranging multiple images of the humidity status monitoring areas inside the file cabinet to obtain a sequence of image blocks of the humidity status monitoring areas inside the file cabinet, the analysis and monitoring of the humidity status changes inside the file cabinet can be further strengthened. By arranging multiple monitoring area images into a sequence, it is convenient for the system to continuously monitor and compare and analyze the humidity status of different areas in the file cabinet in a chronological or spatial order. Arranging the image block sequence also helps to improve the efficiency of data processing and analysis. The system can apply the same or different humidity analysis algorithms to each image block, such as humidity perception or model reasoning based on image processing technology. By processing the image blocks one by one, the system can detect the pattern or trend of humidity changes more quickly and take necessary management measures in time. By analyzing the humidity characteristics of different image blocks in the sequence, the system can establish a comprehensive model of the humidity status inside the file cabinet, including daily changes, seasonal changes, and possible abnormal conditions. This model can provide the system with predictive humidity management suggestions to help optimize file protection strategies and operating procedures.

[0036] In a specific embodiment of the present application, the file cabinet internal state feature encoding subunit 1212 includes: linearly embedding the file cabinet internal humidity state monitoring area image block sequence through the file cabinet internal humidity state monitoring area image block sequence encoder to obtain a sequence of file cabinet internal humidity state monitoring area image block embedding vectors; passing the sequence of file cabinet internal humidity state monitoring area image block embedding vectors through the file cabinet internal state monitoring area image block semantic encoding model based on the ViT model to obtain the file cabinet internal state monitoring area associated feature vector.

[0037] It should be understood that the image block sequence of the humidity status monitoring area inside the file cabinet is a sequence of multiple image blocks, each of which represents the humidity status of different areas in the file cabinet. These image blocks are usually preprocessed and feature extracted, but still exist in the form of images and need to be further converted into numerical vectors so that the computer can understand and process them. Through the linear embedding encoder, each image block can be converted into an embedding vector. Linear embedding is usually implemented using a set of linear mappings to map the high-dimensional image feature space to a low-dimensional vector space. This embedding vector not only retains the important information of the original image block, but also reduces the complexity and storage requirements of the data, while improving the efficiency of subsequent analysis and calculation.

[0038] Furthermore, the sequence of embedding vectors of image blocks in the file cabinet internal humidity status monitoring area is passed through the file cabinet internal status monitoring area image block semantic encoding model based on the ViT (Vision Transformer) model, aiming to further extract and encode the semantic information between image blocks to obtain higher-level file cabinet internal status monitoring related feature vectors. Among them, the ViT model, as an advanced visual processing model, can effectively process and understand the semantic information in the image. Compared with the traditional convolutional neural network, the ViT model uses a self-attention mechanism to capture the global relationship between image blocks, so as to better express the semantic content of the image. This ability enables the ViT model to identify and encode key humidity-related features when processing the file cabinet internal humidity status monitoring image, such as the pattern of humidity distribution and the correlation between regions. The file cabinet internal status monitoring area image block semantic encoding model based on the ViT model takes the embedding vector sequence of each image block as input, and integrates the local features of each image block into a global semantic expression through a multi-layer self-attention mechanism and a fully connected network. The associated feature vector of the file cabinet internal status monitoring area obtained in this way can more accurately reflect the changes and trends of the overall humidity status of the file cabinet. Through the semantic encoding of the ViT model, semantic understanding and contextual perception of the image blocks in the file cabinet internal status monitoring area can be achieved. The model not only considers the features of a single image block, but also comprehensively considers the spatial relationship and semantic association between different image blocks, such as how humidity changes in adjacent areas affect the overall file cabinet humidity state. This comprehensive consideration capability helps the system to more accurately identify potential humidity problem areas and predict humidity change trends in advance. Specifically, the sequence of embedding vectors of the image blocks in the file cabinet internal humidity status monitoring area is input into the converter module of the image blocks semantic encoding model of the file cabinet internal status monitoring area based on the ViT model to obtain the associated feature vector of the file cabinet internal status monitoring area.

[0039] In a specific embodiment of the present application, the filing cabinet environment parameter feature extraction unit 122 includes: passing the environment parameters collected by the environment integrated sensor through the filing cabinet internal environment parameter semantic encoder to obtain multiple filing cabinet environment parameter feature vectors; arranging the multiple filing cabinet environment parameter feature vectors into the filing cabinet environment parameter feature vector.

[0040] It should be understood that the environmental integrated sensor can collect a variety of environmental parameters inside the file cabinet in real time, such as temperature, humidity, air quality, etc. These parameters are crucial for the long-term preservation of files, and therefore require effective management and monitoring. The sensor generates a large amount of environmental data streams through continuous data collection. As a key component for processing environmental data, the semantic encoder of the internal environmental parameters of the file cabinet is responsible for converting the original environmental parameter data into a higher-level, more abstract feature vector representation. This encoding process is usually combined with machine learning techniques, such as deep neural networks or self-attention mechanisms, to capture the complex relationships and patterns between environmental parameters. Specifically, the environmental parameters collected by the environmental integrated sensor are segmented to obtain a word sequence; the embedding layer of the semantic encoder of the internal environmental parameters of the file cabinet is used to map each word in the word sequence to a word embedding vector to obtain a sequence of word embedding vectors; the converter-based Bert model of the semantic encoder of the internal environmental parameters of the file cabinet is used to perform global context semantic encoding on the sequence of the embedding vectors to obtain multiple file cabinet environmental parameter feature vectors.

[0041] Furthermore, arranging the feature vectors of the environmental parameters of the filing cabinet in a sequence form enables the system to continuously track and analyze the changes in environmental parameters in chronological order. Each feature vector represents the comprehensive state of the internal environment of the filing cabinet at a specific time point or time period. By arranging them into a sequence, the system can establish time series data, analyze the changing trends of environmental parameters such as temperature, humidity, air quality, etc., identify periodic changes or abnormal conditions, and thus adjust the environmental control strategy in a timely manner. The serialized feature vectors of the environmental parameters of the filing cabinet can also support the analysis and comparison of spatial relationships. The interior of the filing cabinet is usually divided into multiple areas or levels, and each area may be affected by different environmental conditions. By arranging the feature vector sequence, the system can compare the environmental parameters of different areas, identify the differences and commonalities between the areas, and further optimize the accuracy and efficiency of environmental control.

[0042] In the above-mentioned intelligent file cabinet state monitoring system 100 based on deep learning, the file cabinet dehumidification judgment module 130 is used to judge whether it is necessary to start the file cabinet dehumidification operation based on the associated feature vector of the file cabinet internal state monitoring area and the file cabinet environmental parameter feature vector. It should be understood that the associated feature vector of the file cabinet internal state monitoring area reflects the humidity state and its change trend of different areas or image blocks. These vectors are generated by a deep learning model (such as a semantic encoding model based on ViT), which can capture the spatial distribution and change pattern of humidity in each area of ​​the file cabinet. By analyzing these feature vectors, the system can identify possible abnormal humidity areas, such as local high humidity or areas with rapid changes in humidity. The file cabinet environmental parameter feature vector contains various environmental parameters inside the file cabinet, such as temperature, humidity, air quality, etc. These vectors are usually calculated based on sensor data and reflect the change trend and state of the overall environment of the file cabinet. By monitoring the changes in the environmental parameter feature vector, the system can analyze whether the environmental conditions are close to or have reached the critical value that causes humidity damage to the file. Comprehensive analysis of the associated feature vector of the file cabinet internal state monitoring area and the file cabinet environmental parameter feature vector can make intelligent decisions.

[0043] Figure 4 FIG. 1 is a block diagram of a file cabinet dehumidification judgment module in a smart file cabinet status monitoring system based on deep learning according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, the file cabinet dehumidification judgment module 130 includes: a file cabinet humidity feature fusion unit 131, which is used to fuse the associated feature vector of the file cabinet internal state monitoring area and the file cabinet environmental parameter feature vector to obtain a file cabinet humidity judgment classification feature vector; a file cabinet humidity feature optimization unit 132, which is used to perform class harmony adaptive matching on the file cabinet humidity judgment classification feature vector based on local distribution consistency to obtain an optimized file cabinet humidity judgment classification feature vector; a file cabinet dehumidification operation classification judgment unit 133, which is used to pass the optimized file cabinet humidity judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether it is necessary to start the file cabinet dehumidification operation.

[0044] It should be understood that the fusion of the associated feature vectors of the file cabinet internal state monitoring area and the feature vectors of the file cabinet environmental parameters to obtain the classification feature vectors for file cabinet humidity judgment is intended to comprehensively utilize the local humidity distribution information and the overall environmental parameters to make more accurate and detailed classification judgments on the humidity conditions of the file cabinet. The associated feature vectors of the file cabinet internal state monitoring area reflect the humidity distribution and change trends of different areas in the file cabinet. These vectors are generated by visual processing models (such as semantic encoding models based on ViT) and can capture the humidity state of each area, including possible high humidity areas or areas with abnormal humidity changes. The feature vectors of file cabinet environmental parameters cover various environmental parameters inside the file cabinet, such as temperature, humidity, air quality, etc. These feature vectors are usually calculated based on sensor data and reflect the dynamic changes of the overall environment of the file cabinet. By monitoring the feature vectors of environmental parameters, the system can analyze whether the overall environmental conditions are conducive to the long-term preservation of files, or whether there is a risk of excessive humidity.

[0045] Furthermore, in the technical solution of the present application, it is considered that the humidity state inside the file cabinet is affected by a variety of environmental factors, such as temperature, air circulation, etc. Different file cabinets may be under different environmental conditions, resulting in the diversity of the collected images and environmental parameters, thereby affecting the local distribution of the feature vector. The installation position, angle, resolution, etc. of the camera may be different, resulting in differences in the quality and content of the collected images, thereby affecting the accuracy of regional segmentation and feature extraction. The linear embedding encoder and the semantic encoding model based on the ViT model may not be able to fully capture all relevant feature information, or have different sensitivities to certain features, resulting in differences in the distribution of feature vectors in local areas, thereby resulting in local distribution differences in the fused file cabinet humidity judgment classification feature vector. The file cabinet humidity judgment classification feature vector has local distribution differences, resulting in a class harmony offset relative to the predetermined class label of the classifier when it is classified by the classifier, affecting the classification iteration effect. Based on this, the file cabinet humidity judgment classification feature vector is adaptively matched based on the local distribution consistency.

[0046] The file cabinet humidity judgment classification feature vector is adaptively matched based on the local distribution consistency to obtain the optimized file cabinet humidity judgment classification feature vector, including: determining the weight learning parameter matrix of the classifier; extracting each column vector in the weight learning parameter matrix to obtain a set of weight learning parameter column vectors; using the following optimization formula to respectively calculate the metric correlation prediction factor between each weight learning parameter column vector in the set of weight learning parameter column vectors and the file cabinet humidity judgment classification feature vector to obtain a metric correlation prediction factor modulation vector, wherein the optimization formula is:

[0047]

[0048] Among them, V i represents each weight learning parameter column vector in the set of weight learning parameter column vectors, V c represents the classification feature vector for judging the humidity of the filing cabinet, It means subtracting by position. is the superposition norm of the two norms of the eigenvectors, ||·|| F is the Frobenius norm of the eigenvector, M p represents the weight learning parameter matrix, exp(·) represents the exponential operation with the natural constant e as the base, V p Representing the metric correlation prediction factor modulation vector; using the metric correlation prediction factor modulation vector as a weight vector, calculating the position point multiplication between the metric correlation prediction factor modulation vector and the file cabinet humidity judgment classification feature vector to obtain the optimized file cabinet humidity judgment classification feature vector.

[0049] In the technical solution of the present application, it is considered that there are local distribution differences in the classification feature vectors of the file cabinet humidity judgment, which leads to the problem that when it is classified by the classifier, there will be a class harmony offset relative to the predetermined class label of the classifier, affecting the classification iteration effect. Based on this, in the technical solution of the present application, the classification feature vectors of the file cabinet humidity judgment are adaptively matched based on the local distribution consistency of the class harmony, and the superposition paradigm of the two norms between each column vector in the weight learning parameter matrix of the classifier and the classification feature vectors of the file cabinet humidity judgment is used as the subclass label related metric integral factor and the matrix norm of the weight learning parameter of the classifier is used as the differential factor, and then the association paradigm between the subclass label related metric integral factor and the differential factor is used to perform bidirectional linear manifold modulation on the classification feature vectors of the file cabinet humidity judgment, so as to improve the local feature distribution consistency of the classification feature vectors of the file cabinet humidity judgment by adaptively stretching and compressing the classification feature vectors of the file cabinet humidity judgment, thereby reducing the harmony offset of the classification feature vectors of the file cabinet humidity judgment relative to the predetermined class label of the classifier, thereby improving the accuracy of the classification results.

[0050] Furthermore, the optimized file cabinet humidity judgment classification feature vector contains the comprehensively considered internal humidity status information of the file cabinet. These vectors are obtained through the aforementioned comprehensive analysis, reflecting the humidity distribution in different areas, the changing trend of environmental parameters, and possible humidity anomalies. This information is crucial for accurately judging the current humidity status of the file cabinet. Here, by processing the optimized file cabinet humidity judgment classification feature vector through the classifier, the complex humidity status information can be converted into a concise and clear classification result. Through real-time monitoring and rapid feedback from the classifier, the system can implement timely and accurate environmental protection measures. It avoids delayed response or omissions that may occur during manual monitoring, and improves the efficiency and reliability of environmental management. In addition, automated dehumidification operations can also save energy and resource usage, and achieve the dual goals of environmental protection and long-term preservation of documents through intelligent management methods.

[0051] In summary, the embodiment of the present application first obtains the internal status monitoring image of the file cabinet collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the two. Finally, a classifier is used to determine whether it is necessary to start the file cabinet dehumidification operation, so as to timely discover and deal with potential humidity problems, effectively protect files from moisture and damage, and thus improve the efficiency and quality of file management.

[0052] As described above, the smart file cabinet status monitoring system 100 based on deep learning according to the embodiment of the present application can be implemented in various terminal devices. In one example, the smart file cabinet status monitoring system 100 based on deep learning can be integrated into the terminal device as a software module and / or a hardware module. For example, the smart file cabinet status monitoring system 100 based on deep learning can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the smart file cabinet status monitoring system 100 based on deep learning can also be one of the many hardware modules of the terminal device.

[0053] Alternatively, in another example, the deep learning-based intelligent file cabinet status monitoring system 100 and the terminal device may also be separate devices, and the deep learning-based intelligent file cabinet status monitoring system 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0054] Figure 5 Flowchart of a method for monitoring the state of an intelligent file cabinet based on deep learning according to an embodiment of the present application. Figure 5As shown, according to the deep learning-based intelligent file cabinet status monitoring method of the embodiment of the present application, it includes: S110, obtaining the file cabinet internal status monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environment integrated sensor; S120, extracting the file cabinet internal status monitoring area associated feature vector and the file cabinet environmental parameter feature vector from the file cabinet internal status monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environment integrated sensor; S130, based on the file cabinet internal status monitoring area associated feature vector and the file cabinet environmental parameter feature vector, judging whether it is necessary to start the file cabinet dehumidification operation.

[0055] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned intelligent file cabinet status monitoring method based on deep learning have been referred to above. Figures 1 to 4 The description of the deep learning-based intelligent file cabinet status monitoring system has been introduced in detail, and therefore, its repeated description will be omitted.

[0056] Below, reference Figure 6 To describe an electronic device according to an embodiment of the present application.

[0057] Figure 6 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0058] The processor 11 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solution provided in the embodiment of the present application; the memory 12 can be implemented in the form of a read-only memory (Read Only Memory, ROM), a static storage device, a dynamic storage device or a random access memory (Random Access Memory, RAM), etc.

[0059] The memory 12 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 12, and the processor 11 calls and executes the intelligent file cabinet status monitoring method based on deep learning in the embodiment of this application;

[0060] Input / output interface 13, used to implement information input and output;

[0061] The communication interface 14 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0062] A bus 15 that transmits information between the various components of the device (e.g., the processor 11, the memory 12, the input / output interface 13, and the communication interface 14);

[0063] The processor 11 , the memory 12 , the input / output interface 13 and the communication interface 14 are connected to each other in communication within the device via a bus 15 .

[0064] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium. The storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned deep learning-based smart file cabinet status monitoring method is implemented.

[0065] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0066] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0067] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0068] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0069] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0070] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0071] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0072] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0073] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0074] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method of each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs. The preferred embodiment of the embodiment of the present application is described above with reference to the accompanying drawings, and the scope of rights of the embodiment of the present application is not limited thereto. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall be within the scope of rights of the embodiments of the present application.

Claims

1. A smart file cabinet status monitoring system based on deep learning, characterized in that: include: A humidity status data acquisition module, used to acquire the internal status monitoring image of the filing cabinet collected by the camera installed inside the filing cabinet and the environmental parameters collected by the environmental integrated sensor; A humidity state data extraction module, used to extract a file cabinet internal state monitoring area associated feature vector and a file cabinet environmental parameter feature vector from the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor; The file cabinet dehumidification judgment module is used to judge whether it is necessary to start the file cabinet dehumidification operation based on the associated feature vector of the file cabinet internal state monitoring area and the file cabinet environmental parameter feature vector.

2. The deep learning-based intelligent file cabinet status monitoring system according to claim 1 is characterized in that: The humidity status data extraction module comprises: A file cabinet internal state feature extraction unit, used for extracting features from the file cabinet internal state monitoring image captured by the camera installed inside the file cabinet to obtain a feature vector associated with the file cabinet internal state monitoring area; The file cabinet environment parameter feature extraction unit is used to extract features from the environment parameters collected by the environment integrated sensor to obtain the file cabinet environment parameter feature vector.

3. The deep learning-based intelligent file cabinet status monitoring system according to claim 2 is characterized in that: The file cabinet internal state feature extraction unit comprises: A file cabinet internal humidity state preprocessing subunit, used for performing data preprocessing on the file cabinet internal state monitoring image acquired by the camera installed inside the file cabinet to obtain a file cabinet internal humidity state monitoring area image block sequence; The file cabinet internal state feature encoding subunit is used to perform feature encoding on the image block sequence of the file cabinet internal humidity state monitoring area to obtain the associated feature vector of the file cabinet internal state monitoring area.

4. The deep learning-based intelligent file cabinet status monitoring system according to claim 3 is characterized in that: The file cabinet internal humidity state preprocessing subunit comprises: Performing region segmentation on the file cabinet internal state monitoring image captured by the camera installed inside the file cabinet to obtain a plurality of file cabinet internal humidity state monitoring region images; The plurality of images of the file cabinet internal humidity status monitoring area are arranged to obtain a sequence of image blocks of the file cabinet internal humidity status monitoring area.

5. The deep learning-based intelligent file cabinet status monitoring system according to claim 4 is characterized in that: The file cabinet internal state feature encoding subunit includes: The image block sequence of the internal humidity status monitoring area of ​​the file cabinet is linearly embedded into an encoder to obtain a sequence of embedding vectors of the image blocks of the internal humidity status monitoring area of ​​the file cabinet; The sequence of embedding vectors of the image blocks in the humidity status monitoring area of ​​the file cabinet is passed through the semantic encoding model of the image blocks in the internal status monitoring area of ​​the file cabinet based on the ViT model to obtain the associated feature vector of the internal status monitoring area of ​​the file cabinet.

6. The deep learning-based intelligent file cabinet status monitoring system according to claim 5 is characterized in that: The file cabinet environment parameter feature extraction unit comprises: The environmental parameters collected by the environmental integrated sensor are passed through an internal environmental parameter semantic encoder of the file cabinet to obtain a plurality of file cabinet environmental parameter feature vectors; The plurality of file cabinet environment parameter feature vectors are arranged into the file cabinet environment parameter feature vector.

7. The deep learning-based intelligent file cabinet status monitoring system according to claim 6 is characterized in that: The file cabinet dehumidification judgment module includes: A file cabinet humidity feature fusion unit, used to fuse the file cabinet internal state monitoring area associated feature vector and the file cabinet environmental parameter feature vector to obtain a file cabinet humidity judgment classification feature vector; A file cabinet humidity feature optimization unit, used for performing a class harmony adaptive matching based on local distribution consistency on the file cabinet humidity judgment classification feature vector to obtain an optimized file cabinet humidity judgment classification feature vector; The file cabinet dehumidification operation classification judgment unit is used to pass the optimized file cabinet humidity judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to judge whether the file cabinet dehumidification operation needs to be started.

8. The deep learning-based intelligent file cabinet status monitoring system according to claim 7 is characterized in that: The file cabinet humidity characteristic optimization unit comprises: Determining a weight learning parameter matrix of the classifier; Extracting each column vector in the weight learning parameter matrix to obtain a set of weight learning parameter column vectors; Respectively calculating the metric correlation prediction factor between each weight learning parameter column vector in the set of weight learning parameter column vectors and the file cabinet humidity judgment classification feature vector to obtain a metric correlation prediction factor modulation vector; The metric correlation prediction factor modulation vector is used as a weight vector, and the position point multiplication between the metric correlation prediction factor modulation vector and the file cabinet humidity judgment classification feature vector is calculated to obtain the optimized file cabinet humidity judgment classification feature vector.

9. A method for monitoring the status of an intelligent file cabinet based on deep learning, characterized in that: include: Acquire the internal state monitoring image of the filing cabinet collected by the camera installed inside the filing cabinet and the environmental parameters collected by the environmental integrated sensor; Extracting a file cabinet internal state monitoring area associated feature vector and a file cabinet environmental parameter feature vector from the file cabinet internal state monitoring image acquired by the camera installed inside the file cabinet and the environmental parameters acquired by the environmental integrated sensor; Based on the associated feature vector of the file cabinet internal state monitoring area and the feature vector of the file cabinet environmental parameters, it is determined whether the file cabinet dehumidification operation needs to be started.

10. The method for monitoring the state of an intelligent file cabinet based on deep learning according to claim 9, characterized in that: Extracting the associated feature vector of the internal state monitoring area of ​​the file cabinet and the feature vector of the file cabinet environment parameters from the file cabinet internal state monitoring image collected by the camera installed inside the file cabinet and the environmental parameters collected by the environmental integrated sensor, including: Performing feature extraction on the file cabinet internal state monitoring image captured by the camera installed inside the file cabinet to obtain a feature vector associated with the file cabinet internal state monitoring area; Feature extraction is performed on the environmental parameters collected by the environmental integrated sensor to obtain the file cabinet environmental parameter feature vector.

Images