An intelligent door lock control method, device, equipment and medium

Abstract

The present invention relates to the field of intelligent devices, and its purpose is to provide an intelligent door lock control method, device, equipment and medium. The method includes: acquiring video image information of a target object; extracting a plurality of motion feature vectors based on the video image information; performing clustering analysis on the motion feature vectors to obtain a feature sequence; matching the feature sequence with a preset sequence library, and controlling the door lock to open when the match is successful. By extracting features from the video image information, constructing feature vectors and then performing identification, the present invention can effectively reduce the complexity of identification, improve the accuracy and precision of identification, and effectively improve the security of the intelligent door lock.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent devices, and particularly to an intelligent door lock control method, device, equipment and medium. Background Art

[0002] The brief constituent elements of a door lock system include a lock body and a key. The traditional unlocking mechanism relies on a physical key carried by synthetic metal, plastic, etc. This traditional unlocking mechanism has relatively large security risks and hidden costs. For example, the users of physical keys are anonymous, and the transfer and loss of keys will pose threats to privacy and property. The traditional unlocking with a physical key takes a long time, and there are problems such as the lock core getting stuck and the key breaking, which will greatly affect the user experience.

[0003] The upgraded intelligent door lock based on this has solved some of the above problems, but there are also some defects in the solution. As a biometric technology, face recognition can perform identity authentication by collecting and comparing face features. Although in many occasions, face recognition is widely used as a convenient and efficient identity authentication method, it also has some limitations and problems: when it comes to the application of face recognition technology, two of the main problems are the high false recognition rate and security problems. These problems may lead to adverse effects, such as the system misjudging and being unable to accurately identify the identity, or being deceived by an attacker and exploited system vulnerabilities to steal data. The high false recognition rate is due to the fact that the face recognition system still has limitations in recognizing variable scenarios and face features. For example, in low light conditions, it is difficult to recognize the features of a face. In addition, different facial expressions, hair occlusion, masks, etc. will also affect the accuracy of face recognition. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide an intelligent door lock control method, device, equipment and medium to solve the problems of poor recognition accuracy of human faces and security risks in existing intelligent door locks.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] Embodiments of the present invention provide an intelligent door lock control method, including:

[0007] Obtain video image information of a target object;

[0008] Extract a plurality of motion feature vectors based on the video image information;

[0009] Perform clustering analysis on the motion feature vectors to obtain a feature sequence;

[0010] Match the feature sequence with a preset sequence library, and control the door lock to open when the match is successful.

[0011] Optionally, the extracting of multiple motion feature vectors based on the video image information includes:

[0012] Performing identification and screening on the video image information to obtain multiple motion feature images;

[0013] Extracting feature vectors from the motion feature images to obtain multiple motion feature vectors.

[0014] Optionally, the performing identification and screening on the video image information to obtain motion feature images includes:

[0015] Selecting multiple video frame images from the video image information at a preset time interval;

[0016] Performing identification and analysis on each video frame image to obtain pixel differences;

[0017] Comparing the pixel differences with a preset difference threshold;

[0018] When the pixel difference is greater than the preset difference threshold, extracting the video frame image corresponding to the pixel difference as a motion feature image.

[0019] Optionally, the extracting feature vectors from the motion feature images to obtain multiple motion feature vectors includes:

[0020] Extracting multiple feature vectors from each motion feature image according to preset feature requirements;

[0021] Combining the feature vectors of each motion feature image to obtain multiple feature matrices;

[0022] Performing normalization processing on the feature matrix corresponding to each motion feature image to obtain multiple motion feature vectors.

[0023] Optionally, before performing normalization processing on the feature matrix corresponding to each motion feature image, the method further includes:

[0024] Comparing the dimension of each feature matrix with a preset dimension threshold;

[0025] When the dimension is greater than the preset dimension threshold, performing dimensionality reduction processing on the feature matrix.

[0026] Optionally, the performing clustering analysis on the motion feature vectors to obtain a feature sequence includes:

[0027] Inputting the motion feature vectors into a preset clustering model to obtain multiple feature sets;

[0028] Analyze each of the said feature sets to obtain multiple cluster centers;

[0029] Extract the time information corresponding to each cluster center from the video image information based on the video frame images corresponding to the cluster centers;

[0030] Sort the center vectors corresponding to the cluster centers according to the time information to obtain a feature sequence.

[0031] Optionally, the method further includes:

[0032] Regularly update the preset sequence library according to the successfully matched feature sequence and a preset prediction model.

[0033] An embodiment of the present invention further provides an intelligent door lock control device, including:

[0034] An acquisition module, configured to acquire video image information of a target object;

[0035] An extraction module, configured to extract a plurality of motion feature vectors based on the video image information;

[0036] An analysis module, configured to perform clustering analysis on the motion feature vectors to obtain a feature sequence;

[0037] A control module, configured to match the feature sequence with a preset sequence library, and control the door lock to open when the match is successful.

[0038] An embodiment of the present invention further provides an electronic device, including:

[0039] A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the intelligent door lock control method provided by the embodiment of the present invention.

[0040] An embodiment of the present invention further provides a computer-readable storage medium, which stores computer instructions for causing a computer to execute the intelligent door lock control method provided by the embodiment of the present invention.

[0041] The technical solution of the present invention has the following advantages:

[0042] The present invention provides an intelligent door lock control method, which obtains video image information of a target object; extracts a plurality of motion feature vectors based on the video image information; performs clustering analysis on the motion feature vectors to obtain a feature sequence; and matches the feature sequence with a preset sequence library, and controls the door lock to open when the matching is successful. By extracting features from the video image information, constructing feature vectors and then performing recognition, the present invention can effectively reduce the complexity of recognition, improve the accuracy and precision of recognition, and effectively improve the security of the intelligent door lock. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0044] Figure 1 It is a flowchart of the intelligent door lock control method in the embodiment of the present invention;

[0045] Figure 2 It is a flowchart of extracting a plurality of motion feature vectors based on the video image information in the embodiment of the present invention;

[0046] Figure 3 It is a flowchart of identifying and screening the video image information in the embodiment of the present invention;

[0047] Figure 4 It is a flowchart of extracting feature vectors from the motion feature image in the embodiment of the present invention;

[0048] Figure 5 It is a flowchart of performing dimensionality reduction processing on the feature matrix in the embodiment of the present invention;

[0049] Figure 6 It is a flowchart of performing clustering analysis on the motion feature vectors in the embodiment of the present invention;

[0050] Figure 7 It is a schematic structural diagram of the intelligent door lock control device in the embodiment of the present invention;

[0051] Figure 8 It is a schematic structural diagram of the electronic device in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0053] According to an embodiment of the present invention, an embodiment of an intelligent door lock control method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0054] In this embodiment, an intelligent door lock control method is provided, which can be used for the above terminal devices, such as a computer, etc. Figure 1 As shown, the intelligent door lock control method includes the following steps:

[0055] Step S1: Obtain video image information of a target object. Specifically, the video image information includes the video frame image of each video frame and the corresponding time.

[0056] Step S2: Extract a plurality of motion feature vectors based on the video image information. Specifically, analyze multiple motion types such as the basic features, motion amplitude, and motion intensity of the unlocking action behavior to obtain the motion feature vectors.

[0057] Step S3: Perform clustering analysis on the motion feature vectors to obtain a feature sequence. Specifically, optimize the extraction of multiple motion feature vectors from the video image information through clustering analysis and establish a feature sequence for subsequent matching.

[0058] Step S4: Match the feature sequence with a preset sequence library. When the match is successful, control the door lock to open. Specifically, the preset sequence library stores the feature sequences of multiple target objects input by the user. When the match is successful, open the door for the target object. If the match fails, count the number of failures. When the statistical count exceeds the warning value, an alarm message can be sent to the bound object of the intelligent door lock.

[0059] Through the above steps S1 to S4, the intelligent door lock control method provided by the embodiment of the present invention obtains the video image information of the target object; extracts a plurality of motion feature vectors based on the video image information; performs clustering analysis on the motion feature vectors to obtain a feature sequence; and matches the feature sequence with a preset sequence library, and controls the door lock to open when the match is successful. The present invention extracts features from the video image information, constructs feature vectors and then performs recognition, which can effectively reduce the complexity of recognition, improve the accuracy and precision of recognition, and effectively improve the security of the intelligent door lock.

[0060] Specifically, in one embodiment, step S2 above, as Figure 2 shown, specifically includes the following steps:

[0061] Step S21: Identify and screen the video image information to obtain a plurality of motion feature images. Specifically, when performing identification and screening, multiple motion types such as the basic features, motion amplitude, and motion intensity of the unlocking action behavior can be analyzed according to the actual situation.

[0062] Step S22: Extract feature vectors from the motion feature images to obtain a plurality of motion feature vectors. Specifically, by extracting the motion feature vectors, it is convenient to accurately analyze the subsequent motion situation and determine whether it is consistent with the action sequence stored in the preset sequence library.

[0063] Specifically, in one embodiment, step S21 above, as Figure 3 shown, specifically includes the following steps:

[0064] Step S211: Select a plurality of video frame images from the video image information at a preset time interval. Specifically, the preset time interval can be set according to the required accuracy. The smaller the time interval, the larger the amount of data to be analyzed later, and the higher the accuracy of the analysis result.

[0065] Step S212: Perform identification and analysis on each video frame image to obtain a pixel difference. Specifically, when performing identification and screening, multiple motion types such as the basic features, motion amplitude, and motion intensity of the unlocking action behavior can be analyzed according to the actual situation.

[0066] Step S213: Compare the pixel difference with a preset difference threshold.

[0067] Step S214: When the pixel difference is greater than the preset difference threshold, extract the video frame image corresponding to the pixel difference as a motion feature image.

[0068] Specifically, the accuracy of the modeling and simulation of the background image directly affects the detection effect. Regardless of any moving object detection algorithm, it is necessary to meet the processing requirements of any image scene as much as possible. However, due to the complexity and unpredictability of the scene, as well as the existence of various environmental interferences and noises, such as sudden changes in illumination, fluctuations of some objects in the actual background image, camera jitter, and the impact of moving objects entering and leaving the scene on the original scene, it makes the modeling and simulation of the background relatively difficult. When analyzing and recognizing each video frame image, the background difference method can be used to detect moving objects, which is fast, accurate in detection, and easy to implement.

[0069] Specifically, in one embodiment, the above step S22 is as Figure 4 shown and specifically includes the following steps:

[0070] Step S221: Extract multiple feature vectors from each motion feature image according to preset feature requirements. Specifically, a feature extraction algorithm suitable for the motion feature image can be selected according to needs to extract the feature vectors, such as color histograms, texture features, shape features, etc.

[0071] Step S222: Combine the feature vectors of each motion feature image to obtain multiple feature matrices. Specifically, each feature vector can be used as a row or a column of the matrix, or combined in other ways.

[0072] Step S223: Perform normalization processing on the feature matrix corresponding to each motion feature image to obtain multiple motion feature vectors. Specifically, the normalization processing process can adopt mean-variance normalization or range normalization. Perform normalization processing on the feature matrix to eliminate the dimensional differences between different features.

[0073] Specifically, in one embodiment, before the above step S223, as Figure 5 shown, it specifically includes the following steps:

[0074] Step S22301: Compare the dimension of each feature matrix with a preset dimension threshold.

[0075] Step S22302: When the dimension is greater than the preset dimension threshold, perform dimensionality reduction processing on the feature matrix.

[0076] Specifically, the dimensionality reduction processing process can adopt principal component analysis (PCA) or linear discriminant analysis (LDA) to reduce the feature matrix to a lower dimension to reduce the complexity and computational amount of the data.

[0077] Specifically, in one embodiment, the above step S3, asFigure 6 As shown in Figure 6 , it specifically includes the following steps:

[0078] Step S31: Input the motion feature vector into a preset clustering model to obtain multiple feature sets. Specifically, the preset clustering model can classify the motion feature vectors by category and integrate the motion feature vectors belonging to the same category into the same feature set.

[0079] Step S32: Analyze each feature set to obtain multiple cluster centers. Specifically, the K-means clustering algorithm can be used to select the cluster center of each feature set. By iteratively moving the cluster center until it no longer changes, the cluster center of each feature set can be obtained simply and quickly.

[0080] Step S33: Extract the time information corresponding to each cluster center from the video image information based on the video frame image corresponding to the cluster center.

[0081] Step S34: Sort the center vectors corresponding to the cluster centers according to the time information to obtain a feature sequence. Specifically, sorting according to the time corresponding to each video frame image can effectively ensure the continuity of the sequence, thereby ensuring the coherence and accuracy of action analysis.

[0082] Specifically, in one embodiment, the above intelligent door lock control method further specifically includes the following steps:

[0083] Regularly update the preset sequence library according to the successfully matched feature sequence and the preset prediction model.

[0084] Specifically, the preset prediction model is established in the following manner:

[0085] (1) Construct multiple recurrent layers. Each recurrent layer has the same weight parameters, and each recurrent layer receives the input at the current time step and the hidden state at the previous time step, and outputs the hidden state at the current time step.

[0086] (2) For each time step, calculate the hidden state at the current time step according to the input and the hidden state at the previous time step. It can be expressed by the following formula:

[0087] h_t = f(W_hh * h_{t - 1}+W_xh * x_t + b_h)

[0088] Where h_t is the hidden state at the current time step, h_{t - 1} is the hidden state at the previous time step, x_t is the input at the current time step, W_hh and W_xh are weight matrices, b_h is a bias vector, and f is an activation function (which can be tanh or ReLU).

[0089] (3) Calculate the output based on the hidden state of the current time step. Depending on the task, different methods can be selected to calculate the output. For example, for classification tasks, a fully connected layer can be used to map the hidden state to class labels. For sequences of different motion types, they are divided into sets under different class labels.

[0090] (4) Calculate the loss based on the predicted results output by the model and the real sequence, and use the backpropagation algorithm to update the network parameters to minimize the loss function. The optimization algorithms used here include but are not limited to Stochastic Gradient Descent (SGD), Adam.

[0091] (5) Train the RNN using the training data, and improve the model performance by iteratively optimizing the network parameters. After training is completed, the trained model can be used to analyze the successfully matched feature sequences, classify the generated new sequence data, and update it to the preset sequence database regularly. Thus, when facing different images or expression actions of the target object, it can perform recognition faster and more accurately, improving the user experience while having higher security.

[0092] Specifically, in one embodiment, the above intelligent door lock control method further includes the following steps:

[0093] Count the data matching frequencies of each target object in the preset sequence database at preset date intervals, and delete the feature sequences with data matching frequencies of each target object lower than the preset frequency. Specifically, in this way, the data storage volume can be reduced, and not only the space utilization rate can be improved but also the matching efficiency can be improved while meeting the security requirements, bringing a better use experience to users.

[0094] In this embodiment, an intelligent door lock control device is also provided. This device is used to implement the above embodiments and preferred implementation manners, and the descriptions that have been given will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0095] This embodiment provides an intelligent door lock control device, as Figure 7 shown, including:

[0096] An acquisition module 101, configured to acquire video image information of a target object. For detailed content, refer to the relevant description of step S1 in the above method embodiment, and details will not be repeated here.

[0097] An extraction module 102, configured to extract multiple motion feature vectors based on the video image information. For detailed content, refer to the relevant description of step S2 in the above method embodiment, and details will not be repeated here.

[0098] An analysis module 103 is configured to perform clustering analysis on the motion feature vectors to obtain a feature sequence. For detailed content, refer to the relevant description of step S3 in the above method embodiment, which will not be elaborated here.

[0099] A control module 104 is configured to match the feature sequence with a preset sequence library, and control the door lock to open when the match is successful. For detailed content, refer to the relevant description of step S4 in the above method embodiment, which will not be elaborated here.

[0100] The intelligent door lock control device in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0101] The further function descriptions of the above-mentioned respective modules are the same as those in the corresponding above embodiments, and will not be elaborated here.

[0102] According to an embodiment of the present invention, there is also provided an electronic device, as Figure 8 shown. The electronic device may include a processor 801 and a memory 802. The processor 801 and the memory 802 may be connected through a bus or other means. Figure 8 Taking the connection through the bus as an example.

[0103] The processor 801 may be a central processing unit (CPU). The processor 801 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. chips, or combinations of the above types of chips.

[0104] The memory 802, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the method in the method embodiment of the present invention. The processor 801 executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions, and modules stored in the memory 802, that is, implements the method in the above method embodiment.

[0105] The memory 802 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created by the processor 801 and the like. In addition, the memory 802 may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory 802 may optionally include a memory remotely provided with respect to the processor 801, and these remote memories can be connected to the processor 801 through 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.

[0106] One or more modules are stored in the memory 802 and, when executed by the processor 801, execute the methods in the above-mentioned method embodiments.

[0107] For specific details of the above-mentioned electronic device, reference can be made to the corresponding relevant descriptions and effects in the above-mentioned method embodiments for understanding, and details will not be elaborated here.

[0108] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned method embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memories.

[0109] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. An intelligent door lock control method, characterized in that, Including: Obtain the video image information of the target object; Extract multiple motion feature vectors based on the video image information; Perform clustering analysis on the motion feature vectors to obtain a feature sequence; Match the feature sequence with a preset sequence library, and control the door lock to open when the match is successful; The performing clustering analysis on the motion feature vectors to obtain a feature sequence includes: Input the motion feature vectors into a preset clustering model to obtain multiple feature sets; Analyze each of the feature sets to obtain multiple clustering centers and the center vectors corresponding to each clustering center; Extract the time information corresponding to each clustering center from the video image information based on the video frame image corresponding to the clustering center; Sort the center vectors corresponding to the clustering centers according to the time information to obtain a feature sequence.

2. The intelligent door lock control method according to claim 1, characterized in that The extracting multiple motion feature vectors based on the video image information includes: Perform recognition and screening on the video image information to obtain multiple motion feature images; Extract feature vectors from the motion feature images to obtain multiple motion feature vectors.

3. The intelligent door lock control method according to claim 2, wherein The performing recognition and screening on the video image information to obtain motion feature images includes: Select multiple video frame images from the video image information at a preset time interval; Perform recognition and analysis on each of the video frame images to obtain pixel differences; Compare the pixel differences with a preset difference threshold; When the pixel difference is greater than the preset difference threshold, extract the video frame image corresponding to the pixel difference as a motion feature image.

4. The intelligent door lock control method according to claim 2, wherein The extracting feature vectors from the motion feature images to obtain multiple motion feature vectors includes: Extract multiple feature vectors from each of the motion feature images according to preset feature requirements; Combine the feature vectors of each of the motion feature images to obtain multiple feature matrices; Perform normalization processing on the feature matrix corresponding to each of the motion feature images to obtain multiple motion feature vectors.

5. The intelligent door lock control method according to claim 4, characterized in that, Before performing normalization processing on the feature matrix corresponding to each of the motion feature images, the method further includes: Compare the dimension of each of the feature matrices with a preset dimension threshold; When the dimension is greater than the preset dimension threshold, perform dimensionality reduction processing on the feature matrix.

6. The intelligent door lock control method according to claim 1, wherein, The method further includes: Regularly update the preset sequence library according to the feature sequence with successful match and a preset analysis model.

7. An intelligent door lock control device, characterized in that, Including: An obtaining module, configured to obtain the video image information of the target object; An extracting module, configured to extract multiple motion feature vectors based on the video image information; An analyzing module, configured to perform clustering analysis on the motion feature vectors to obtain a feature sequence; A control module, configured to match the feature sequence with a preset sequence library, and control the door lock to open when the match is successful; Specifically, the analysis module is configured to input the motion feature vector into a preset clustering model to obtain multiple feature sets; analyze each of the feature sets to obtain multiple cluster centers and a center vector corresponding to each cluster center; extract time information corresponding to each cluster center from the video image information based on the video frame image corresponding to the cluster center; and sort the center vectors corresponding to the cluster centers according to the time information to obtain a feature sequence.

8. An electronic device, characterized in that, Comprising: A memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to execute the intelligent door lock control method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to execute the intelligent door lock control method according to any one of claims 1-6.

Images