An equipment ledger management system based on convolutional neural network model

Abstract

The present invention discloses an equipment ledger management system based on a convolutional neural network model, and relates to the technical field of electronic ledgers. The system includes: an equipment image acquisition module, an equipment image processing module, and an equipment ledger data management module; wherein the equipment image acquisition module is used to acquire images acquired by a remote image acquisition device; the equipment image processing module is used to perform image recognition on the acquired images to determine the cabinet information and host information in the images; the cabinet information includes the cabinet number, and the host information includes the location, model, manufacturer information, internal accessory information, and the size of the space occupied by the cabinet; the equipment ledger data management module is used to acquire and store the cabinet information and host configuration information. In this way, efficient image recognition and automatic information update are achieved, and the accuracy and efficiency of equipment information management in the computer room are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic ledgers, and in particular to an equipment ledger management system based on a convolutional neural network model. Background Art

[0002] In the related art, for cross-regional enterprises, the enterprise's computer rooms are distributed in different locations, and these computer rooms are purchased, deployed and managed by local branches. The headquarters has limited knowledge of the host equipment configuration of these computer rooms, and due to equipment damage, upgrades, or replacement of accessories, changes in equipment configuration information often occur, making it difficult for the headquarters to accurately understand the actual configuration of the equipment. In the related art, the enterprise's equipment ledger system is used to number and manage the cabinets and host equipment in computer rooms in various places.

[0003] The equipment inventory system records detailed information about the host equipment, including equipment model, configuration, etc. However, host equipment information is usually entered manually, such as through previous purchase lists. Manual entry may result in errors or delays, which leads to differences between the data in the system and the actual configuration of computer rooms in various locations. Summary of the invention

[0004] The present invention provides an equipment inventory management system based on a convolutional neural network model to solve the technical problem that errors or entry delays may occur when entering data manually such as purchase lists, resulting in differences between data in the system and actual configurations.

[0005] According to one aspect of the present disclosure, a device inventory management system based on a convolutional neural network model is provided, the system comprising: a device image acquisition module, a device image processing module, and a device inventory data management module;

[0006] Wherein, the device image acquisition module is used to acquire the device image acquired by the remote image acquisition device;

[0007] The device image processing module is used to perform image recognition on the collected image to determine the cabinet information and host information in the image; the cabinet information includes the cabinet number, and the host information includes the location, model, manufacturer information, internal accessories information and the size of the space occupied by the cabinet;

[0008] The equipment inventory data management module is used to obtain and store the cabinet information and host configuration information.

[0009] In a possible implementation, the device image processing module is used to perform image recognition on the collected image to determine the cabinet information and host information in the image, including: the device image processing module uses a convolutional neural network to perform image recognition on the collected image.

[0010] In a possible implementation, the device image processing module is also used to train a convolutional neural network model:

[0011] adding a label to the acquired image, wherein the label indicates the type of equipment or the number of the cabinet;

[0012] Preprocessing the collected image data;

[0013] Choose a suitable convolutional neural network model;

[0014] Dividing the acquired images into a training set, a validation set and a test set, and using the training data to perform model training on the image data in the training set;

[0015] Using the validation set to validate the performance parameters of the model, and adjusting and optimizing the model according to the performance parameters;

[0016] Use the test set to evaluate the performance of the model.

[0017] In a possible implementation, the device image acquisition module is used to acquire images acquired by a remote image acquisition device, including: the remote image acquisition device includes a mobile terminal or a fixed camera device in a computer room.

[0018] In a possible implementation, the preprocessing of the acquired image data includes: performing size standardization, color correction and noise removal on the acquired image;

[0019] The using the training data to perform model training on the image data in the training set includes: using the training set data to train the CNN model, updating the weights and parameters of the model through a back propagation algorithm to minimize the loss function, and using data enhancement technology to expand the training data to improve the generalization ability of the model, wherein the data enhancement technology includes rotation, flipping, and scaling;

[0020] The performance parameters include: accuracy, recall, and precision;

[0021] The adjusting and optimizing the model includes: adjusting the hyperparameters and learning rate of the model.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] Different from the current mainstream convolutional neural network systems developed based on Tensorflow or Pytorch libraries, the former is more complex and not suitable for small and medium-sized enterprises to write according to their own needs. The latter requires experienced algorithm engineers to tune. Both are costly for this system in terms of model availability, deployment convenience and ecosystem. Therefore, we implemented a simplified convolutional neural network model based on the actual application situation.

[0024] An equipment inventory management system based on a convolutional neural network model according to an embodiment of the present disclosure uses an equipment image acquisition module to acquire images acquired by a remote image acquisition device. The convolutional neural network performs image recognition on the acquired images to determine the cabinet information in the image, including the cabinet number, the location, model, manufacturer information of the host, internal accessories information, or host configuration information of the size of the space occupied by the cabinet. The equipment inventory data management module is used to acquire and store the cabinet information and host information. For example, the equipment inventory data management module stores the cabinet information and host information in a database, or automatically updates historical information based on the acquired cabinet information and host information. In this way, efficient image recognition and automatic information update are achieved, thereby improving the accuracy and efficiency of computer room equipment information management. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A block diagram of an equipment inventory management system based on a convolutional neural network model according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0026] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0027] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0028] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present disclosure.

[0029] Figure 1 A block diagram of a device inventory management system based on a convolutional neural network model according to an embodiment of the present disclosure is shown. Figure 1As shown, the system includes: a device image acquisition module, a device image processing module, and a device ledger data management module;

[0030] Wherein, the device image acquisition module is used to acquire images acquired by a remote image acquisition device;

[0031] The device image processing module is used to perform image recognition on the collected image to determine the cabinet information and host information in the image; the cabinet information includes the cabinet number, and the host information includes the location, model, manufacturer information, internal accessory information, or the size of the space occupied by the cabinet;

[0032] The equipment inventory data management module is used to obtain and store the cabinet information and host information.

[0033] The equipment image acquisition module is used to acquire images captured by a remote image acquisition device, including: the remote image acquisition device includes a mobile terminal or a fixed camera device in the computer room. For example, the staff at the computer room site uses a mobile device to take images of the equipment in the cabinet, and uploads the images to the equipment ledger management system. The neural network model in the equipment ledger management system will return the results of image recognition and update the equipment information in the ledger system based on the results. Among them, the image recognition results include the number information of the cabinet and the location, model, manufacturer information of the equipment in the cabinet, or the size of the space occupied by the cabinet. Alternatively, the fixed camera device in the computer room captures images of the cabinets or equipment in the computer room, and uploads them to the equipment ledger management system for image recognition and data update.

[0034] For example, for cross-regional enterprises, computer rooms are spread all over the place, and each computer room has a unique code to identify its location. In each computer room, the cabinets have specific codes to identify the locations of different cabinets. The equipment ledger management system based on the convolutional neural network model takes photos of the equipment in the cabinet, and then uses the AI ​​algorithm for image recognition to automatically identify the cabinet number and the location, model, manufacturer information and U number of the host equipment, thereby determining the specific information of each device in the ledger system. So that the on-site staff can compare the information in the ledger system with the real configuration information of the host, so that the enterprise can grasp the real situation of the enterprise equipment in a timely manner. Among them, the U number is the size of the space occupied by the equipment in the cabinet. Some servers are thicker and occupy 2U, which means that they occupy two standard cabinet positions in the cabinet. A cabinet usually has 42U, that is, 42 units of space.

[0035] If there are 10 hard disks in the cabinet as a disk array, and one disk is broken, the replacement disk may not be of the same model. For example, if the power module is broken, one module is replaced, and the other one is not replaced yet, the equipment configuration list will be different from the original one. At this time, the staff on site in the computer room need to manually enter the update status of the equipment in the computer room, which is not only labor-intensive and error-prone, but also may not be entered in time.

[0036] Traditional computer vision techniques and rule-based methods are used for image recognition. These methods usually include the following techniques:

[0037] Feature extraction and traditional machine learning: Previous methods usually involve manually designing image features (such as color, texture, edges, etc.), and then using traditional machine learning algorithms (such as support vector machines, decision trees, etc.) to train classifiers. These methods require a lot of feature engineering for specific image recognition tasks and are not flexible enough to adapt to complex and diverse scenarios.

[0038] Rule-based image processing: In some applications, image recognition is based on predefined rules and pattern matching. These rules usually need to be set manually and are only applicable to specific situations. This approach is limited in flexibility and is not suitable for tasks that need to adapt to changes.

[0039] In a possible implementation, the device image processing module is used to perform image recognition on the collected image to determine the cabinet information and host information in the image, including: the device image processing module uses a convolutional neural network to perform image recognition on the collected image.

[0040] For example, the collected image is subjected to image recognition by a convolutional neural network to determine the cabinet information in the image, including the cabinet number, the location, model, manufacturer information of the host, or the host information of the space occupied by the cabinet.

[0041] The equipment inventory data management module is used to obtain and store the cabinet information and host information. For example, the equipment inventory data management module stores the cabinet information and host information in a database, or automatically updates historical information based on the obtained cabinet information and host information. In this way, efficient image recognition and automatic information update are achieved, thereby improving the accuracy and efficiency of equipment information management in the computer room.

[0042] In a possible implementation, the device image processing module is also used to train a convolutional neural network model:

[0043] adding a label to the acquired image, wherein the label indicates the type of equipment or the number of the cabinet;

[0044] Preprocessing the collected image data;

[0045] Choose a suitable convolutional neural network model;

[0046] Dividing the acquired images into a training set, a validation set and a test set, and using the training data to perform model training on the image data in the training set;

[0047] Using the validation set to validate the performance parameters of the model, and adjusting and optimizing the model according to the performance parameters;

[0048] Use the test set to evaluate the performance of the model.

[0049] For example, data collection and preparation include: collecting image datasets containing equipment in cabinets, which include images of various equipment, angles, and lighting conditions. Labeling images, that is, assigning corresponding labels to each image to indicate the type of each device, cabinet number, and other key information. Preprocessing image data, including image size standardization, color correction, denoising, etc., to ensure the consistency and quality of input data.

[0050] Image recognition model selection and architecture design: Choose an appropriate CNN architecture, such as VGG, ResNet, Inception, etc., or customize a model. Define the number of layers and structure of the model, including convolutional layers, pooling layers, fully connected layers, etc., to adapt to the complexity of the task.

[0051] Data splitting: Divide the dataset into training, validation, and test sets, usually using common ratios such as 70% training, 15% validation, and 15% testing.

[0052] Model compilation: After selecting a model, you need to compile the model and define the loss function, optimizer, and evaluation indicators. The loss function usually uses the cross entropy loss function, the optimizer can choose Adam, SGD, etc., and the evaluation indicators can choose accuracy, precision, recall, etc.

[0053] Model training: Input the prepared training set into the model for training. You can set parameters such as the number of training rounds and batch size. After each round of training, you can use the validation set to evaluate the performance of the model to avoid overfitting of the model.

[0054] Model testing: After training is completed, the test set is used to evaluate the performance of the model, and indicators such as the accuracy, precision, and recall of the model on the test set can be obtained.

[0055] Model optimization: If the model performs poorly, you can improve the performance of the model by adjusting hyperparameters, increasing the amount of data, using data enhancement, transfer learning, etc.

[0056] Model saving and deployment: After model training is completed, the model can be saved for subsequent deployment and use. When deploying, it can be deployed using servers, cloud, mobile terminals, etc.

[0057] It should be noted that during the model training process, operations such as data set annotation, data set expansion, data set division, and hyperparameter debugging are also required. At the same time, hardware acceleration devices such as GPU and TPU can also be used to increase the speed of model training.

[0058] In a possible implementation, the device image acquisition module is used to acquire images acquired by a remote image acquisition device, including: the remote image acquisition device includes a mobile terminal or a fixed camera device in a computer room.

[0059] In a possible implementation, the preprocessing of the acquired image data includes: performing size standardization, color correction and noise removal on the acquired image;

[0060] The using the training data to perform model training on the image data in the training set includes: using the training set data to train the CNN model, updating the weights and parameters of the model by a back propagation algorithm to minimize the loss function so that the model can accurately recognize the image, and using data enhancement technology to expand the training data to improve the generalization ability of the model, wherein the data enhancement technology includes rotation, flipping, and scaling;

[0061] Among them, weight adjustment refers to the process of updating and optimizing the weight parameters in the network during the neural network training process. The purpose of adjusting the weight is to enable the neural network to better approximate the true distribution of the training data, thereby achieving better generalization ability.

[0062] Weight adjustment is achieved through the back-propagation algorithm, which determines the update direction and size of network parameters by calculating the error between the network output and the true label.

[0063] When using the cross entropy loss function, the loss function formula is as follows:

[0064] ;

[0065] Where N is the number of samples, is the true label of the sample, is the predicted value output by the network.

[0066] In each round of training, it is necessary to calculate the gradient of the loss function with respect to all weight parameters in the network, that is, ;

[0067] is the weight parameter of the jth neuron in the i-th layer of the network. Then the gradient descent algorithm is used to update the weight parameter:

[0068] ;

[0069] in, is the learning rate, which is used to control the update step size of the weight parameters.

[0070] In practical applications, we use batch normalization technology to accelerate network convergence and improve training stability; we can use momentum method and adaptive learning rate technology to further improve training efficiency and accuracy. In addition, we can also use regularization technology to control the overfitting phenomenon of the network and improve generalization ability.

[0071] In order to compare the optimization effect, we use indicators such as precision, recall, and F1-score to observe the optimization effect.

[0072] Precision: The proportion of samples that the model predicts to be positive samples that are actually positive samples.

[0073] Recall rate: The proportion of samples that are actually positive samples that are predicted by the model to be positive samples.

[0074] F1-score: An indicator that comprehensively considers precision and recall, which can be regarded as the harmonic mean of precision and recall.

[0075] After optimization, for the classification task, the precision and recall rates both reached over 80%, the F1-score was over 0.8, and the loss function value range was below 0.1 or below 0.01. The model was preliminarily considered qualified.

[0076] Model evaluation: Use the test data set to evaluate the performance of the model, including accuracy, recall, F1 value and other indicators. If the model performs poorly, you can consider adjusting hyperparameters, optimizing model structure, increasing the amount of data and other methods to improve performance.

[0077] Model deployment: Deploy the trained model to actual applications, which can be done using servers, cloud, mobile terminals, etc.

[0078] When implementing convolutional neural networks, you can use deep learning frameworks such as TensorFlow, PyTorch, Keras, etc. to accelerate model training and optimization. At the same time, you can also combine hardware acceleration devices such as GPU and TPU to increase the training and prediction speed of the model.

[0079] In practical applications, the performance and generalization ability of the model can be further improved through data enhancement, transfer learning, model fusion, etc.

[0080] The performance parameters include: accuracy, recall, and precision;

[0081] The adjusting and optimizing the model includes: adjusting the hyperparameters and learning rate of the model.

[0082] Using the Convolutional Neural Network (CNN) model for image recognition brings the following advancements:

[0083] Automatic feature learning: CNN models can automatically learn features from data without manually designing complex feature extraction processes, which greatly reduces the workload of feature engineering.

[0084] High Accuracy: CNN models perform well in image classification tasks and can achieve high accuracy. They are able to handle larger and more complex datasets and are sensitive to subtle differences.

[0085] Generalization ability: CNN models can generalize to different data sets and diverse scenarios, making them more suitable for various applications, including image recognition, object detection, image segmentation, etc.

[0086] Transferability: Already trained CNN models (such as pre-trained neural networks) can be used for multiple different tasks, thus speeding up the development process.

[0087] Real-time performance and real-time feedback: CNN models can be used in real-time or near-real-time applications to provide users with fast image recognition results.

[0088] In short, the use of convolutional neural networks for image recognition has higher automation, accuracy, and applicability than previous traditional methods. Especially in the processing of large-scale and complex image data sets, the CNN model shows obvious advancement.

[0089] Different from the current mainstream convolutional neural network systems developed based on Tensorflow or Pytorch libraries, the former is more complex and not suitable for small and medium-sized enterprises to write according to their own needs. The latter requires experienced algorithm engineers to tune. Both are costly for this system in terms of model availability, deployment convenience and ecosystem. Therefore, we implemented a simplified convolutional neural network model based on the actual application situation.

[0090] An equipment inventory management system based on a convolutional neural network model according to an embodiment of the present disclosure uses an equipment image acquisition module to acquire images acquired by a remote image acquisition device. The convolutional neural network performs image recognition on the acquired images to determine the cabinet information in the image, including the cabinet number, the location, model, manufacturer information of the host, or information about the size of the space occupied by the host in the cabinet. The equipment inventory data management module is used to acquire and store the cabinet information and host information. For example, the equipment inventory data management module stores the cabinet information and host information in a database, or automatically updates historical information based on the acquired cabinet information and host information. In this way, efficient image recognition and automatic information updating are achieved, thereby improving the accuracy and efficiency of computer room equipment information management.

[0091] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present disclosure is not limited by the order of the actions described, because according to the present disclosure, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present disclosure.

[0092] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0093] In the several embodiments provided in the present disclosure, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the units, which 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, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0094] The units described 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.

[0095] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software program module.

[0096] If the integrated unit is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present disclosure, 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, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc., various media that can store program codes.

[0097] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable memory, which may include a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0098] The embodiments of the present disclosure are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present disclosure. The description of the above embodiments is only used to help understand the method of the present disclosure and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present disclosure, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present disclosure.

[0099] Various aspects of the present disclosure are described herein with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0100] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0101] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A equipment ledger management system based on a convolutional neural network model, characterized in that: The system includes: an equipment image acquisition module, an equipment image processing module, and an equipment ledger data management module; Wherein, the device image acquisition module is used to acquire the device image acquired by the remote image acquisition device; The device image processing module is used to perform image recognition on the collected image to determine the cabinet information and host information in the image; the cabinet information includes the cabinet number, and the host information includes the location, model, manufacturer information, internal accessories information and the size of the space occupied by the cabinet; The equipment inventory data management module is used to obtain and store the cabinet information and host configuration information; The device image processing module is used to perform image recognition on the collected image to determine the cabinet information and host information in the image, including: the device image processing module uses a convolutional neural network to perform image recognition on the collected image; The device image processing module is also used to train the convolutional neural network model: adding a label to the acquired image, wherein the label indicates the type of equipment or the number of the cabinet; Preprocessing the collected image data; Choose a suitable convolutional neural network model; Dividing the collected images into a training set, a validation set and a test set, and using the training data to perform model training on the image data in the training set; Using the validation set to validate the performance parameters of the model, and adjusting and optimizing the model according to the performance parameters; Use the test set to evaluate the performance of the model.

2. According to claim 1, a device ledger management system based on a convolutional neural network model is characterized in that: The device image acquisition module is used to acquire images acquired by a remote image acquisition device, including: the remote image acquisition device includes a mobile terminal or a fixed camera device in a computer room.

3. According to claim 1, a device ledger management system based on a convolutional neural network model is characterized in that: The preprocessing of the collected image data includes: size standardization, color correction and noise removal of the collected image; The using the training data to perform model training on the image data in the training set includes: using the training set data to train the CNN model, updating the weights and parameters of the model through a back propagation algorithm to minimize the loss function, and using data enhancement technology to expand the training data to improve the generalization ability of the model, wherein the data enhancement technology includes rotation, flipping, and scaling; The performance parameters include: accuracy, recall, and precision; The adjusting and optimizing the model includes: adjusting the hyperparameters and learning rate of the model.

Images