Occupancy identification method, device, equipment and storage medium in smart station

By using the placeholder recognition system in a smart station, combined with image and wireless recognition modules, the problem of low efficiency in private parking space recognition management is solved, and intelligent recognition and accuracy of vehicle placeholder behavior is achieved.

CN118570770BActive Publication Date: 2024-11-22SHENZHEN DIANDIAN ELECTRIC NETWORK TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411046147.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-01
Publication Date
2024-11-22
Estimated Expiration
2044-08-01

AI Technical Summary

Technical Problem

The problem of intelligently identifying the space occupancy of private parking spaces in the prior art has resulted in low efficiency and difficulty in placeholder identification management.

Method used

The placeholder recognition system is adopted in the smart station, combining image recognition and wireless recognition modules, and the image and wireless recognition module are activated when the vehicle is approached by the wake-up module, and timing images and wireless tag information are collected to perform vehicle entry and placeholder recognition, and to judge Whether to start the alarm module.

Benefits of technology

It realizes intelligent identification of vehicle placeholding behavior, improves identification accuracy and efficiency, and improves the intelligence and efficiency of placeholding management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118570770B_ABST
    Figure CN118570770B_ABST
Patent Text Reader

Abstract

The present invention relates to a method, device, equipment and storage medium for identifying occupancy in a smart parking lot, and to the field of parking lot management technology, including: when a vehicle approaches a target parking space, waking up an image recognition module and a wireless recognition module, collecting time-series images, performing vehicle entry recognition, and obtaining a vehicle entry recognition result; when it is a vehicle entering, performing vehicle recognition on the time-series images to obtain an image vehicle recognition result; identifying the wireless tag of the entering vehicle through the wireless recognition module to obtain a wireless vehicle recognition result; according to the image vehicle recognition result and the wireless vehicle recognition result, performing a judgment to obtain an occupancy recognition result, and judging whether to start an alarm module to alarm or not according to the occupancy recognition result. The present invention solves the problem of unintelligent and cumbersome parking space occupancy management in the prior art, and achieves the technical effect of improving the intelligence, accuracy and convenience of parking space management.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of station management, and in particular to a method, device, equipment and storage medium for identifying a placeholder in a smart station. Background Art

[0002] At present, some parking spaces in parking lots are private parking spaces for fixed car owners, and some parking spaces are public parking spaces. However, private parking spaces are often occupied. In order to solve this problem, ground locks (parking space locks) are installed on parking spaces in the prior art. However, the ground locks require car owners to open the ground locks every time they leave, which is very cumbersome to operate. The identification of parking space occupancy often requires inspections by station managers, which is inefficient. Therefore, the prior art cannot intelligently identify the occupancy of private parking spaces, and there are technical problems such as low efficiency and difficulty in occupancy identification management. Summary of the invention

[0003] The present invention aims to solve the problem that private parking spaces cannot be intelligently identified in the prior art, and the technical problems of low efficiency and great difficulty in occupancy identification management, and provides an occupancy identification method, device, equipment and storage medium in a smart station.

[0004] In a first aspect, the present invention provides a method for identifying a placeholder in a smart station, which is applied to a placeholder identification system. The system includes a wake-up module, an image recognition module, a wireless recognition module and an alarm module. The method includes: when a vehicle approaches a target parking space, the wake-up module is started, and the image recognition module and the wireless recognition module are woken up; the image recognition module is used to collect time-series images, perform vehicle entry identification, and obtain a vehicle entry identification result; when the vehicle entry identification result is vehicle entry, the time-series images are used to perform vehicle identification to obtain an image vehicle identification result, and the wireless identification module is used to identify the wireless tag of the entering vehicle to obtain a wireless vehicle identification result; based on the image vehicle identification result and the wireless vehicle identification result, a judgment is made to obtain a placeholder identification result, and based on the placeholder identification result, it is determined whether to start the alarm module to alarm or not.

[0005] In a second aspect, the present invention provides a placeholder identification device in a smart station, which is connected to a placeholder identification system. The system includes a wake-up module, an image recognition module, a wireless recognition module and an alarm module. The device includes: a judgment wake-up unit, which is used to start the wake-up module and wake up the image recognition module and the wireless recognition module when the vehicle approaches the target parking space; a vehicle entry identification unit, which is used to collect time-series images through the image recognition module, perform vehicle entry identification, and obtain a vehicle entry identification result; a vehicle identification unit, which is used to perform vehicle identification on the time-series images when the vehicle entry identification result is vehicle entry, obtain an image vehicle identification result, and identify the wireless tag of the entering vehicle through the wireless identification module to obtain a wireless vehicle identification result; an alarm judgment unit, which is used to judge and obtain a placeholder identification result based on the image vehicle identification result and the wireless vehicle identification result, and judge whether to start the alarm module to alarm or not based on the placeholder identification result.

[0006] In a third aspect, the present invention provides an electronic device comprising: a memory for storing a computer software program; and a processor for reading and executing the computer software program, thereby implementing the occupancy identification method in the smart station of the first aspect.

[0007] In a fourth aspect, the present invention provides a non-transitory computer-readable storage medium, in which a computer software program is stored. When the computer software program is executed by a processor, the method for occupancy identification in a smart station of the first aspect is implemented.

[0008] The present invention starts the wake-up module when the vehicle approaches the target parking space, and wakes up the image recognition module and the wireless recognition module, collects time-series images, performs vehicle entry recognition, obtains vehicle entry recognition results, performs vehicle recognition on the time-series images when the vehicle enters, obtains image vehicle recognition results, recognizes the wireless tag of the entering vehicle through the wireless recognition module, obtains wireless vehicle recognition results, makes a judgment based on the image vehicle recognition results and the wireless vehicle recognition results to obtain the occupancy recognition results, and judges whether to start the alarm module to alarm or not based on the occupancy recognition results. The solution provided by the present invention can intelligently identify vehicle occupancy behavior, and can improve the accuracy of vehicle recognition through dual recognition of images and wireless tags. It is suitable for different scenarios, realizes vehicle occupancy recognition in parking lots, and achieves the technical effect of improving the accuracy and efficiency of occupancy recognition, as well as improving the efficiency and intelligence of occupancy management. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A schematic diagram of the flow of the method for identifying a placeholder in a smart station provided by the present invention;

[0010] Figure 2A schematic diagram of the structure of the occupancy identification device in the smart station provided by the present invention;

[0011] Figure 3 A functional block diagram of the electronic device provided by the present invention;

[0012] Figure 4 A functional block diagram of a computer-readable storage device provided by the present invention.

[0013] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0014] The wake-up determination unit 100 , the vehicle entry recognition unit 200 , the vehicle recognition unit 300 , the alarm determination unit 400 , the electronic device 500 , the memory 510 , the processor 520 , the computer program 511 , the computer-readable storage medium 600 , and the computer program 611 . DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0016] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, "plurality" means two or more, unless otherwise clearly and specifically defined.

[0017] In the description of the present application, the term "for example" is used to mean "used as an example, illustration or description". Any embodiment described as "for example" in the present application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid unnecessary details to obscure the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present application.

[0018] Example 1

[0019] like Figure 1 As shown, an embodiment of the present invention provides a method for identifying a placeholder in a smart station, which is applied to a placeholder identification system. The system includes a wake-up module, an image recognition module, a wireless recognition module and an alarm module, including:

[0020] S10: When the vehicle approaches the target parking space, the wake-up module is started, and the image recognition module and the wireless recognition module are woken up;

[0021] In the embodiment of the present application, the target parking space is a private parking space to be identified as occupied. When the owner of the private parking space needs to perform parking space occupancy identification management, the method provided in the embodiment of the present application is implemented.

[0022] When a vehicle approaches a target parking space, the wake-up module installed near the parking space first detects the approach of the vehicle. The wake-up module can achieve this detection function through a variety of sensors, such as infrared sensors, ultrasonic sensors, or geomagnetic sensors. When these sensors sense the presence of a vehicle, the wake-up module immediately sends a signal to start the subsequent recognition module to perform subsequent steps to identify the parking space occupancy.

[0023] First, the wake-up module activates the image recognition module. At the same time, the wake-up module also activates the wireless recognition module. After receiving the start signal, the image recognition module immediately enters the working state and starts to initialize the camera or camera to prepare for image acquisition. The main function of the image recognition module is to analyze the collected time-series images to identify whether the vehicle has entered the target parking space. This recognition usually uses advanced image processing technologies, such as machine vision and deep learning algorithms, to accurately detect the entry of the vehicle.

[0024] After receiving the start signal, the wireless identification module starts to initialize related devices, such as RFID readers or Bluetooth receivers, and prepares to detect and process wireless signals. The main function of the wireless identification module is to identify the wireless tags carried by vehicles entering the parking space. These tags can be RFID tags or Bluetooth devices. By reading the information of these tags, the system can further confirm the identity and specific information of the vehicle.

[0025] This continuous startup process ensures that the system can respond quickly and perform multiple verifications the moment the vehicle approaches the target parking space. By combining image recognition and wireless recognition, the system can effectively improve the accuracy and reliability of recognition, providing reliable data support for subsequent occupancy recognition and alarms.

[0026] Step S10 in the method provided in the embodiment of the present application includes:

[0027] Collecting wake-up feature information when the vehicle approaches the target parking space through one or more of an infrared sensor, a sound sensor, and a vibration sensor in the wake-up module;

[0028] According to the wake-up feature information, wake-up analysis is performed to obtain a wake-up analysis result, and when the wake-up analysis result is yes, the image recognition module and the wireless recognition module are awakened.

[0029] In the embodiment of the present application, one or more of an infrared sensor, a sound sensor, and a vibration sensor are installed in the wake-up module, and it continuously works to monitor whether there is a vehicle approaching the target parking space.

[0030] The infrared sensor senses the approach of a vehicle by detecting infrared radiation or heat signals emitted by the vehicle. The sound sensor determines the approach of a vehicle by capturing the noise of the vehicle engine or the sound signals of the tires rubbing against the ground. The vibration sensor confirms the approach of a vehicle by sensing the vibration signals generated by the ground due to the vehicle's movement. The above signals serve as wake-up feature information.

[0031] After obtaining the sensing signal through one or more sensors, the collected original signal is first filtered and processed to remove environmental noise and irrelevant signals to ensure the accuracy of the data. Then, the processed signal is used as the wake-up feature information to compare and match with the preset feature pattern, such as the preset heat signal, sound signal or vibration signal, to determine whether it meets the characteristics of the approaching vehicle, perform wake-up analysis, and generate a wake-up analysis result.

[0032] The wake-up analysis results include yes and no. If the wake-up analysis result is "yes", that is, it is confirmed that the vehicle is approaching the target parking space, then the wake-up module will send instructions to start the image recognition module and the wireless recognition module. Otherwise, it will not be started and continue to monitor whether the vehicle is approaching.

[0033] The analysis of the wake-up feature information includes, for example, determining whether the temperature signal reaches a temperature threshold, determining whether the sound signal reaches a preset decibel threshold, etc., to determine whether there is a vehicle approaching the target parking space and generate a wake-up analysis result.

[0034] S20: collecting time-series images through the image recognition module, performing vehicle entry recognition, and obtaining a vehicle entry recognition result;

[0035] In an embodiment of the present application, after the image recognition module is started, the camera or camera within the image recognition module begins to continuously capture images of the parking area. These images are arranged in chronological order to form time-series images. These time-series images capture the entire process of the vehicle approaching and entering the parking space, providing continuous visual data.

[0036] Furthermore, image processing techniques and algorithms are used to analyze the collected time-series images, and by analyzing the features in the images, such as the proportion of the vehicle covering the target parking space, it is determined whether the vehicle is entering the target parking space. These technologies may include machine vision, deep learning, edge detection, and motion detection.

[0037] After analysis and processing, the image recognition module generates a vehicle entry recognition result. This result indicates whether the vehicle has successfully entered the target parking space. If the image analysis shows that the vehicle has indeed entered the parking space, the recognition result is "vehicle entered"; if not, the result is "not entered". In this way, after the vehicle enters the parking space, subsequent occupancy recognition is performed. In the case that the vehicle may have parked in the wrong parking space and has not fully entered the target parking space and left, occupancy recognition is not performed, saving computing power.

[0038] Step S20 in the method provided in the embodiment of the present application includes:

[0039] The image recognition module continuously collects images in the target parking space to obtain a time-series image;

[0040] The time series images are recognized through a first deep learning model to obtain a vehicle entry recognition result, wherein the first deep learning model includes a first classifier, a second classifier and a third classifier, which respectively input and recognize the time series images.

[0041] In the embodiment of the present application, after the image recognition module is awakened, it uses a camera to continuously shoot the target parking space to generate time-series images. These images are arranged in chronological order and record the entire process of the vehicle approaching, entering and staying in the parking space.

[0042] The time series images are recognized through the first deep learning model to obtain the vehicle entry recognition result. The first deep learning model includes a first classifier, a second classifier and a third classifier, which respectively input and recognize the time series images. The first classifier, the second classifier and the third classifier are connected in series, and are respectively used to recognize different stages of the vehicle entering the target parking space, such as the stages where the vehicle covers different proportions of the target parking space, so as to recognize the proportion of the vehicle entering the target parking space at different stages, and finally obtain the vehicle entry recognition result of whether the vehicle has completely entered the target parking space.

[0043] The step of "recognizing the time series images through the first deep learning model to obtain a vehicle entry recognition result" in the method provided in the embodiment of the present application includes:

[0044] Input the time series image into the first classifier for recognition, and obtain a first parking space coverage classification result, wherein the training data of the first classifier are parking space photos and photos of parking spaces covered at a first ratio, and the classification recognition labels are parking spaces covered at a first ratio and parking spaces not covered at a first ratio;

[0045] When the first parking space coverage classification result is that the parking space is covered at a first ratio, the time series image is input into the second classifier for recognition to obtain a second parking space coverage classification result, wherein the training data of the second classifier are parking space photos and photos of parking spaces covered at a second ratio, and the classification recognition labels are parking spaces covered at a second ratio and parking spaces not covered at a second ratio;

[0046] When the second parking space coverage classification result is that the parking space is covered at a second ratio, the time series image is input into the third classifier for recognition to obtain a third parking space coverage classification result, wherein the training data of the third classifier are parking space photos and photos of parking spaces covered at a third ratio, and the classification recognition labels are parking spaces covered at a third ratio and parking spaces not covered at a third ratio;

[0047] When the third parking space coverage classification result is that the parking space is covered by the third ratio, a vehicle entry identification result including the vehicle entering is generated, otherwise a vehicle entry identification result including the vehicle not entering is generated, wherein the third ratio is greater than the second ratio, and the second ratio is greater than the first ratio.

[0048] In the embodiment of the present application, the image recognition module continuously collects time-series images with a time sequence, and first inputs the real-time collected time-series images into the first classifier for recognition to classify and determine whether the vehicle enters and covers a first proportion of the target parking space.

[0049] The training data of the first classifier includes two types of photos: ordinary photos of parking spaces and photos of parking spaces covered with a first ratio (smaller ratio). Exemplarily, the first ratio is 20% of the target parking space area.

[0050] The classification identification labels of the first classifier are the first proportion of parking spaces covered and the first proportion of parking spaces not covered. Exemplarily, the first classifier can be constructed by a convolutional neural network, and then ordinary photos of parking spaces and photos of parking spaces covered with a first proportion are collected as training data, and the classification identification labels of the first proportion of parking spaces covered and the first proportion of parking spaces not covered are identified based on the ordinary photos of parking spaces and the photos of parking spaces covered with a first proportion, and then classification training is performed until the training converges to complete the construction of the first classifier.

[0051] When it is determined that the vehicle enters and covers the first proportion of the target parking space, that is, when the first parking space coverage classification result is that the parking space is covered by the first proportion, the time series image is input into the second classifier for further recognition.

[0052] Among them, the training data of the second classifier includes photos of parking spaces and photos of the second coverage ratio (medium ratio) of parking spaces, and the classification identification labels include "the second coverage ratio of parking spaces" and "the second uncovered ratio of parking spaces". The second classifier can be trained using the same training method as the first classifier.

[0053] When the second parking space coverage classification result shows that the parking space is covered by the second proportion, the time series image is input into the third classifier for final recognition.

[0054] Among them, the training data of the third classifier includes photos of parking spaces and photos of parking spaces covered at a third ratio (a larger ratio), and the classification identification labels include "parking spaces covered at a third ratio" and "parking spaces not covered at a third ratio". The third classifier can be trained using the same training method as the first and second classifiers.

[0055] The third ratio is greater than the second ratio, and the second ratio is greater than the first ratio. Exemplarily, the third ratio is 80% and the second ratio is 50%. In this way, combined with the first ratio, the vehicle is identified at different stages of entering the target parking space.

[0056] Furthermore, if the third parking space coverage classification result shows that the parking space is covered by the third proportion, the system generates a vehicle entry identification result of "vehicle entry" and determines whether the vehicle has entered the parking space for subsequent occupancy identification. If the third parking space coverage classification result shows that the parking space is not covered by the third proportion, the system continues to monitor the vehicle entering the parking space until a vehicle entry identification result of "vehicle entry" is obtained.

[0057] Optionally, if the vehicle is continuously monitored and does not enter the target parking space, the first deep learning model can be restarted to identify the time series image for re-identification. If the first parking space coverage classification result is that the parking space is not covered in the first proportion, the recognition of the time series image is terminated.

[0058] Through this step-by-step and precise recognition process in the above steps, the system can accurately determine whether the vehicle enters the target parking space, providing reliable data support for subsequent occupancy recognition and management.

[0059] S30: when the vehicle entry recognition result is vehicle entry, performing vehicle recognition on the time-series image to obtain an image vehicle recognition result, and recognizing the wireless tag of the vehicle entering through the wireless recognition module to obtain a wireless vehicle recognition result;

[0060] In an embodiment of the present application, when it is determined that the vehicle has entered the target parking space and the vehicle entry identification result is generated as "vehicle entry", the system will further identify the vehicle in detail. First, the time-series image is input into the image recognition module again, and the specific features of the vehicle are identified using more sophisticated image processing technology. These features include license plates, models, etc. The system generates an image vehicle recognition result containing detailed vehicle information by analyzing these features. At the same time, the system reads the wireless tag information of the entering vehicle through the wireless identification module. These tags are usually RFID tags or Bluetooth devices installed on the vehicle and contain the unique identification information of the vehicle. After reading this information, the wireless identification module generates a wireless vehicle identification result, which also contains detailed information of the vehicle. By combining the image vehicle identification results and the wireless vehicle identification results, the system can comprehensively and accurately identify the vehicles entering the parking space, providing reliable data support for subsequent management and monitoring.

[0061] Based on the image vehicle recognition results and wireless vehicle recognition results, the vehicle occupancy is judged by scene to improve the intelligence of vehicle occupancy recognition management.

[0062] Step S30 in the method provided in the embodiment of the present application includes:

[0063] Extract the latest N images in the time series image, input them into the second deep learning model, and identify and obtain the license plate recognition result, the license plate presence recognition result and the vehicle model recognition result, wherein the second deep learning model includes a feature extraction network, a time series extraction network, a fourth classifier, a fifth classifier and a sixth classifier;

[0064] When the license plate recognition result, the license plate presence recognition result and the vehicle type recognition result are a preset license plate, a license plate presence and a preset vehicle type, an image vehicle recognition result including the owner's vehicle is generated; otherwise, an image vehicle recognition result of a non-owner's vehicle is generated;

[0065] The wireless identification module identifies the wireless tag entering the vehicle. If a preset wireless tag exists, a wireless vehicle identification result including the authorization of the vehicle owner is generated. If not, a wireless vehicle identification result including the unauthorized authorization of the vehicle owner is generated.

[0066] In an embodiment of the present application, when the vehicle approaches the target parking space and generates a vehicle entry recognition result, the system will further perform detailed vehicle recognition. First, the system extracts the latest N images from the time-series image, for example, extracts the image from the -N frame to the current frame at the current moment, as N images, where N is an integer greater than 1. The N images record the latest status and features of the vehicle. Then, these images are input into the second deep learning model, which consists of a feature extraction network, a time-series extraction network, a fourth classifier, a fifth classifier, and a sixth classifier. The feature extraction network extracts key features from the image, and the time-series extraction network analyzes the time-series features of the image.

[0067] Among them, the fourth classifier performs license plate recognition and generates a license plate recognition result; the fifth classifier determines whether the license plate is in the correct position, specifically whether it is in the correct position on the car, to avoid other people occupying the position by using a fake license plate, and generates a license plate in place recognition result; the sixth classifier performs vehicle model recognition and generates a vehicle model recognition result.

[0068] When the license plate recognition result, the license plate presence recognition result and the vehicle type recognition result are a preset license plate (e.g., the preset license plate of the target parking space owner), the license plate presence and the preset vehicle type, an image vehicle recognition result including the owner's vehicle is generated; otherwise, an image vehicle recognition result of a non-owner's vehicle is generated;

[0069] And, the wireless tag of the vehicle entering the parking space is identified through the wireless identification module. If the preset wireless tag is detected, a wireless vehicle identification result including the vehicle owner's authorization is generated; if the preset wireless tag is not detected, a wireless vehicle identification result including the vehicle owner's unauthorized authorization is generated. Through this continuous identification process, it is possible to accurately identify whether the vehicle entering the parking space is the owner's vehicle and whether it has a wireless tag authorized by the owner, and provide reliable data support for subsequent management and monitoring.

[0070] Exemplarily, the wireless tag is an RFID tag, and the vehicle can be tagged based on the RFID technology in the prior art.

[0071] The step of "extracting the latest N images in the time series image, inputting them into the second deep learning model, and identifying and obtaining the license plate recognition result, the license plate presence recognition result, and the vehicle model recognition result" in the method provided in the embodiment of the present application includes:

[0072] Extract the latest N images in the time series image, input them into the feature extraction network in the second deep learning model, and obtain image feature information, where N is an integer greater than 1;

[0073] Inputting the N images into a time series extraction network in the second deep learning model to obtain time information;

[0074] Input the time information and image feature information into the fourth classifier, and output a license plate recognition result;

[0075] Input the time information and the image feature information into the fifth classifier, and output a license plate presence recognition result, wherein the classification recognition labels of the fifth classifier include no license plate, license plate on the vehicle, and license plate not on the vehicle;

[0076] The time information and the image feature information are input into the sixth classifier, and a vehicle type recognition result is obtained by output, wherein the classification recognition labels of the sixth classifier include SUV, three-box sedan, and two-box sedan.

[0077] In the embodiment of the present application, when the system recognizes that the vehicle enters the target parking space, the latest N images are extracted from the time-series images, where N is an integer greater than 1. These images are input into the feature extraction network in the second deep learning model to obtain image feature information, such as license plate, vehicle model, etc.

[0078] Exemplarily, the feature extraction network is a CNN network, which extracts feature information in N images through convolution of the feature extraction network as a basis for subsequent recognition.

[0079] Input N images into the time series extraction network in the second deep learning model to obtain time information. Exemplarily, the time series extraction network is a GRU or LSTM network, which extracts the time information of N images for subsequent identification of whether the license plate is at a preset position in the time sequence.

[0080] The time information and image feature information are input into the fourth classifier, and the fourth classifier combines these two types of information for processing and outputs the license plate recognition result. The license plate recognition result includes the probability of the recognized license plate number, and the license plate number with the highest probability is selected as the license plate recognition result. The fourth classifier, the fifth classifier, and the sixth classifier are all fully connected networks, which are connected to the feature extraction network and the timing extraction network, and recognize the license plate, vehicle model, and license plate position according to the image feature information and time information.

[0081] The training data of the fourth classifier includes image feature information and time information of vehicle photos, and the classification labels include the probabilities of multiple license plates or no license plate, and license plate recognition is performed through supervised training.

[0082] The training of the fourth classifier adopts, for example, gradient descent or other training methods in the prior art, which will not be described in detail here.

[0083] The time information and image feature information are input into the fifth classifier, and the fifth classifier combines these two types of information for processing and outputs the license plate presence recognition result. The training data of the fifth classifier includes image feature information and time information, and the classification recognition labels include: no license plate, license plate on the car, and license plate not on the car.

[0084] The time information and the image feature information are input into the sixth classifier, and the sixth classifier combines these two types of information for processing and outputs the vehicle type recognition result. The training data of the sixth classifier includes the image feature information and time information of the vehicle image, and the classification recognition labels include: SUV, three-box sedan, two-box sedan. Optionally, it can also include pickup trucks, etc.

[0085] Through the above steps, the vehicle's license plate, license plate status and vehicle model information can be accurately identified as the most important data for determining the occupancy status, providing reliable data support for subsequent management and monitoring. By identifying the license plate status, the situation of duplicate license plate occupying the seat can be avoided, improving the accuracy of occupancy identification.

[0086] S40: according to the image vehicle recognition result and the wireless vehicle recognition result, a determination is made to obtain an occupation recognition result, and according to the occupation recognition result, it is determined whether to start an alarm module to alarm or not.

[0087] In the embodiment of the present application, after completing image vehicle recognition and wireless vehicle recognition, occupancy determination is performed by integrating these recognition results.

[0088] Exemplarily, a comprehensive judgment is made based on the image vehicle recognition result and the wireless vehicle recognition result to obtain the placeholder recognition result. The specific judgment logic may include checking whether the image vehicle recognition result is the owner's vehicle and whether the wireless tag is authorized by the owner. If both the image vehicle recognition result and the wireless vehicle recognition result show that the vehicle information matches and is legal, the system generates a placeholder recognition result of non-placeholder, otherwise the placeholder recognition result including the placeholder is identified and a specific judgment is made.

[0089] Finally, the system determines whether to activate the alarm module based on the occupancy recognition result. If the occupancy recognition result is "occupied", the system activates the alarm module, sends out an alarm signal, notifies the management personnel or takes other measures. If the occupancy recognition result is "not occupied", the system does not activate the alarm module, and the parking space continues to be used normally. Through this coherent judgment and alarm mechanism, the system can effectively manage the use of parking spaces, prevent illegal occupation, and improve the management efficiency and safety of parking lots.

[0090] Step S40 in the method provided in the embodiment of the present application includes:

[0091] When the image vehicle recognition result is the owner's vehicle and the wireless vehicle recognition result is the owner's authorization, it is determined that a non-occupied placeholder recognition result is obtained;

[0092] When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is authorized by the owner, determining the placeholder recognition result of the placeholder authorized by the owner;

[0093] When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is an unauthorized vehicle by the owner, determining and obtaining an occupied place recognition result;

[0094] When the place-occupancy identification result is place-occupancy, it is determined that the alarm module is started to alarm; otherwise, it is determined that the alarm module is not started.

[0095] In the embodiment of the present application, when the image vehicle recognition result is the owner's vehicle and the wireless vehicle recognition result is the owner's authorization, the system determines that a non-occupied parking space recognition result is obtained. This means that the owner's vehicle is legally parked in the parking space, which is normal parking by the owner and will not trigger an alarm.

[0096] When the image vehicle recognition result is a non-owner vehicle and the wireless vehicle recognition result is the owner's authorization, the system determines that the parking space is authorized by the owner. This means that although the vehicle parked in the parking space is not the owner's vehicle, the vehicle is authorized by the owner, such as the authorized vehicle of the owner's family or friends, so the alarm will not be triggered. For example, the owner's authorized RFID tag can be set in the vehicle of the owner's family or friends to realize the owner's authorization tag.

[0097] When the image vehicle recognition result is a non-owner vehicle and the wireless vehicle recognition result is that the owner is not authorized, the system determines that someone else is occupying the target parking space, that is, the occupancy recognition result of the space is occupied. This indicates that the parking space is occupied by an unauthorized vehicle and measures need to be taken.

[0098] Based on the occupancy recognition result, the system determines whether to activate the alarm module. If the occupancy recognition result is occupied, the system activates the alarm module, sends out an alarm signal, notifies the management personnel or takes other measures. If the occupancy recognition result is not occupied or the owner authorizes the parking space to be occupied, the system determines that the alarm module will not be activated and the parking space will continue to be used normally.

[0099] Optionally, when the occupancy recognition result is occupied, a reminder device, such as a speaker, can be arranged in the target parking space, and the availability of public parking spaces in the parking lot can be obtained in combination with the Internet of Things, and voice prompts can be given to encourage the occupant to drive the vehicle to the public parking space for parking.

[0100] Optionally, when the occupancy recognition result is non-occupied, the charging pile in the target parking space can be controlled to start for the car owner to charge.

[0101] The embodiments of the present application achieve the following technical effects through the above specific implementation methods:

[0102] The present invention starts the wake-up module when the vehicle approaches the target parking space, and wakes up the image recognition module and the wireless recognition module, collects time-series images, performs vehicle entry recognition, obtains vehicle entry recognition results, performs vehicle recognition on the time-series images when the vehicle enters, obtains image vehicle recognition results, recognizes the wireless tag of the entering vehicle through the wireless recognition module, obtains wireless vehicle recognition results, makes a judgment based on the image vehicle recognition results and the wireless vehicle recognition results to obtain the occupancy recognition results, and judges whether to start the alarm module to alarm or not based on the occupancy recognition results. The solution provided by the present invention can intelligently identify vehicle occupancy behavior, and can improve the accuracy of vehicle recognition through dual recognition of images and wireless tags. It is suitable for different scenarios, realizes vehicle occupancy recognition in parking lots, and achieves the technical effect of improving the accuracy and efficiency of occupancy recognition, as well as improving the efficiency and intelligence of occupancy management.

[0103] Example 2

[0104] like Figure 2 As shown, based on the same inventive concept as the place occupancy recognition method in the smart station provided in Example 1, an embodiment of the present invention provides a place occupancy recognition device in the smart station, the device is connected to a place occupancy recognition system, the system includes a wake-up module, an image recognition module, a wireless recognition module and an alarm module, the device includes:

[0105] The identification and awakening unit 100 is used to start the awakening module and awaken the image recognition module and the wireless recognition module when the vehicle approaches the target parking space;

[0106] The vehicle entry recognition unit 200 is used to collect time-series images through the image recognition module, perform vehicle entry recognition, and obtain a vehicle entry recognition result;

[0107] The vehicle identification unit 300 is used to perform vehicle identification on the time-series image to obtain an image vehicle identification result when the vehicle entry identification result is a vehicle entry, and to identify the wireless tag of the vehicle entering through the wireless identification module to obtain a wireless vehicle identification result;

[0108] The alarm determination unit 400 is used to determine an occupancy recognition result based on the image vehicle recognition result and the wireless vehicle recognition result, and determine whether to start the alarm module to alarm or not based on the occupancy recognition result.

[0109] Furthermore, the wake-up determination unit 100 is further used for:

[0110] Collecting wake-up feature information when the vehicle approaches the target parking space through one or more of an infrared sensor, a sound sensor, and a vibration sensor in the wake-up module;

[0111] According to the wake-up feature information, wake-up analysis is performed to obtain a wake-up analysis result, and when the wake-up analysis result is yes, the image recognition module and the wireless recognition module are awakened.

[0112] Furthermore, the vehicle entry identification unit 200 is also used for:

[0113] The image recognition module continuously collects images in the target parking space to obtain a time-series image;

[0114] The time series images are recognized through a first deep learning model to obtain a vehicle entry recognition result, wherein the first deep learning model includes a first classifier, a second classifier and a third classifier, which respectively input and recognize the time series images.

[0115] The method of recognizing the time series images through the first deep learning model to obtain a vehicle entry recognition result includes:

[0116] Input the time series image into the first classifier for recognition, and obtain a first parking space coverage classification result, wherein the training data of the first classifier are parking space photos and photos of parking spaces covered at a first ratio, and the classification recognition labels are parking spaces covered at a first ratio and parking spaces not covered at a first ratio;

[0117] When the first parking space coverage classification result is that the parking space is covered at a first ratio, the time series image is input into the second classifier for recognition to obtain a second parking space coverage classification result, wherein the training data of the second classifier are parking space photos and photos of parking spaces covered at a second ratio, and the classification recognition labels are parking spaces covered at a second ratio and parking spaces not covered at a second ratio;

[0118] When the second parking space coverage classification result is that the parking space is covered at a second ratio, the time series image is input into the third classifier for recognition to obtain a third parking space coverage classification result, wherein the training data of the third classifier are parking space photos and photos of parking spaces covered at a third ratio, and the classification recognition labels are parking spaces covered at a third ratio and parking spaces not covered at a third ratio;

[0119] When the third parking space coverage classification result is that the parking space is covered by the third ratio, a vehicle entry identification result including the vehicle entering is generated, otherwise a vehicle entry identification result including the vehicle not entering is generated, wherein the third ratio is greater than the second ratio, and the second ratio is greater than the first ratio.

[0120] Furthermore, the vehicle identification unit 300 is also used for:

[0121] Extract the latest N images in the time series image, input them into the second deep learning model, and identify and obtain the license plate recognition result, the license plate presence recognition result and the vehicle model recognition result, wherein the second deep learning model includes a feature extraction network, a time series extraction network, a fourth classifier, a fifth classifier and a sixth classifier;

[0122] When the license plate recognition result, the license plate presence recognition result and the vehicle type recognition result are a preset license plate, a license plate presence and a preset vehicle type, an image vehicle recognition result including the owner's vehicle is generated; otherwise, an image vehicle recognition result of a non-owner's vehicle is generated;

[0123] The wireless identification module identifies the wireless tag entering the vehicle. If a preset wireless tag exists, a wireless vehicle identification result including the authorization of the vehicle owner is generated. If not, a wireless vehicle identification result including the unauthorized authorization of the vehicle owner is generated.

[0124] The latest N images in the time series image are extracted and input into the second deep learning model to obtain the license plate recognition result, the license plate presence recognition result and the vehicle model recognition result, including:

[0125] Extract the latest N images in the time series image, input them into the feature extraction network in the second deep learning model, and obtain image feature information, where N is an integer greater than 1;

[0126] Inputting the N images into a time series extraction network in the second deep learning model to obtain time information;

[0127] Input the time information and image feature information into the fourth classifier, and output a license plate recognition result;

[0128] Input the time information and the image feature information into the fifth classifier, and output a license plate presence recognition result, wherein the classification recognition labels of the fifth classifier include no license plate, license plate on the vehicle, and license plate not on the vehicle;

[0129] The time information and the image feature information are input into the sixth classifier, and a vehicle type recognition result is obtained by output, wherein the classification recognition labels of the sixth classifier include SUV, three-box sedan, and two-box sedan.

[0130] Furthermore, the alarm determination unit 400 is also used for:

[0131] When the image vehicle recognition result is the owner's vehicle and the wireless vehicle recognition result is the owner's authorization, it is determined that a non-occupied placeholder recognition result is obtained;

[0132] When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is authorized by the owner, determining the placeholder recognition result of the placeholder authorized by the owner;

[0133] When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is an unauthorized vehicle by the owner, determining and obtaining an occupied place recognition result;

[0134] When the place-occupancy identification result is place-occupancy, it is determined that the alarm module is started to alarm; otherwise, it is determined that the alarm module is not started.

[0135] See also Figure 3 , Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 511, the following steps are implemented: when a vehicle approaches a target parking space, the wake-up module is started, and the image recognition module and the wireless recognition module are woken up; the image recognition module is used to collect time-series images, perform vehicle entry recognition, and obtain a vehicle entry recognition result; when the vehicle entry recognition result is vehicle entry, the time-series images are used to perform vehicle recognition, and an image vehicle recognition result is obtained; the wireless recognition module is used to recognize the wireless tag of the vehicle entering, and a wireless vehicle recognition result is obtained; based on the image vehicle recognition result and the wireless vehicle recognition result, a judgment is made to obtain an occupancy recognition result, and based on the occupancy recognition result, it is judged whether to start the alarm module to alarm or not.

[0136] See also Figure 4 , Figure 4 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 4 As shown, the present embodiment provides a computer-readable storage medium 600 on which a computer program 611 is stored. When the computer program 611 is executed by a processor, the following steps are implemented: when a vehicle approaches a target parking space, the wake-up module is started, and the image recognition module and the wireless recognition module are woken up; the image recognition module is used to collect time-series images, perform vehicle entry recognition, and obtain a vehicle entry recognition result; when the vehicle entry recognition result is vehicle entry, the time-series images are used to perform vehicle recognition, and an image vehicle recognition result is obtained; the wireless recognition module is used to recognize the wireless tag of the vehicle entering, and a wireless vehicle recognition result is obtained; based on the image vehicle recognition result and the wireless vehicle recognition result, a judgment is made to obtain an occupancy recognition result, and based on the occupancy recognition result, it is determined whether to start the alarm module to alarm or not.

[0137] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0138] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0139] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0140] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0142] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0143] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for identifying a placeholder in a smart station, characterized in that: The method is applied to a placeholder recognition system, the system comprising a wake-up module, an image recognition module, a wireless recognition module and an alarm module, and the method comprises: When the vehicle approaches the target parking space, the wake-up module is activated and wakes up the image recognition module and the wireless recognition module; The image recognition module collects time-series images, performs vehicle entry recognition, and obtains vehicle entry recognition results; When the vehicle entry recognition result is vehicle entry, performing vehicle recognition on the time-series image to obtain an image vehicle recognition result, and recognizing the wireless tag of the vehicle entering through the wireless recognition module to obtain a wireless vehicle recognition result; According to the image vehicle recognition result and the wireless vehicle recognition result, a determination is made to obtain an occupation recognition result, and according to the occupation recognition result, it is determined whether to start an alarm module to alarm or not; Wherein, when the vehicle entry recognition result is vehicle entry, vehicle recognition is performed on the time-series image to obtain an image vehicle recognition result, and a wireless tag of the vehicle entering is recognized by the wireless recognition module to obtain a wireless vehicle recognition result, including: Extract the latest N images in the time series image, input them into the second deep learning model, and identify and obtain the license plate recognition result, the license plate presence recognition result and the vehicle model recognition result, wherein the second deep learning model includes a feature extraction network, a time series extraction network, a fourth classifier, a fifth classifier and a sixth classifier; When the license plate recognition result, the license plate in position recognition result and the vehicle type recognition result are a preset license plate, a license plate in position and a preset vehicle type, an image vehicle recognition result including the owner's vehicle is generated; otherwise, an image vehicle recognition result of a non-owner's vehicle is generated; Identify the wireless tag entering the vehicle through the wireless identification module, if there is a preset wireless tag, generate a wireless vehicle identification result including the vehicle owner's authorization, if not, generate a wireless vehicle identification result including the vehicle owner's unauthorized; The method of obtaining a placeholder recognition result by distinguishing the image vehicle recognition result and the wireless vehicle recognition result, and determining whether to start the alarm module to alarm or not according to the placeholder recognition result includes: When the image vehicle recognition result is the owner's vehicle and the wireless vehicle recognition result is the owner's authorization, it is determined that a non-occupied placeholder recognition result is obtained; When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is authorized by the owner, determining the placeholder recognition result of the placeholder authorized by the owner; When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is an unauthorized vehicle by the owner, determining and obtaining an occupied place recognition result; When the place-occupancy identification result is place-occupancy, it is determined that the alarm module is started to alarm; otherwise, it is determined that the alarm module is not started.

2. The method for identifying occupancy in a smart station according to claim 1, characterized in that: When the vehicle approaches the target parking space, the wake-up module is activated, including: Collecting wake-up feature information when the vehicle approaches the target parking space through one or more of an infrared sensor, a sound sensor, and a vibration sensor in the wake-up module; According to the wake-up feature information, wake-up analysis is performed to obtain a wake-up analysis result, and when the wake-up analysis result is yes, the image recognition module and the wireless recognition module are awakened.

3. The method for identifying occupancy in a smart station according to claim 1, characterized in that: The image recognition module collects time-series images, performs vehicle entry recognition, and obtains vehicle entry recognition results, including: Continuously collecting images in the target parking space by the image recognition module to obtain the time-series image; The time series images are recognized through a first deep learning model to obtain the vehicle entry recognition result, wherein the first deep learning model includes a first classifier, a second classifier and a third classifier, which respectively input and recognize the time series images.

4. The method for identifying occupancy in a smart station according to claim 3, characterized in that: The time series images are recognized by a first deep learning model to obtain a vehicle entry recognition result, including: Input the time series images into the first classifier for identification, and obtain a first parking space coverage classification result, wherein the training data of the first classifier are parking space photos and photos of parking spaces with a first coverage ratio, and the classification identification labels are parking spaces with a first coverage ratio and parking spaces with a first uncovered ratio; When the first parking space coverage classification result is that the parking space is covered at a first ratio, the time series image is input into the second classifier for recognition to obtain a second parking space coverage classification result, wherein the training data of the second classifier are parking space photos and photos of parking spaces covered at a second ratio, and the classification recognition labels are parking spaces covered at a second ratio and parking spaces not covered at a second ratio; When the second parking space coverage classification result is that the parking space is covered at a second ratio, the time series image is input into the third classifier for recognition to obtain a third parking space coverage classification result, wherein the training data of the third classifier are parking space photos and photos of parking spaces covered at a third ratio, and the classification recognition labels are parking spaces covered at a third ratio and parking spaces not covered at a third ratio; When the third parking space coverage classification result is that the parking space is covered by the third ratio, a vehicle entry identification result including the vehicle entering is generated, otherwise a vehicle entry identification result including the vehicle not entering is generated, wherein the third ratio is greater than the second ratio, and the second ratio is greater than the first ratio.

5. The method for identifying occupancy in a smart station according to claim 1, characterized in that: Extract the latest N images in the time series image, input them into the second deep learning model, and identify and obtain the license plate recognition result, the license plate presence recognition result and the vehicle model recognition result, including: Extract the latest N images in the time series image, input them into the feature extraction network in the second deep learning model, and obtain image feature information, where N is an integer greater than 1; Inputting the N images into the time series extraction network in the second deep learning model to obtain time information; Input the time information and image feature information into the fourth classifier, and output a license plate recognition result; Input the time information and the image feature information into the fifth classifier, and output a license plate presence recognition result, wherein the classification recognition labels of the fifth classifier include no license plate, license plate on the vehicle, and license plate not on the vehicle; The time information and the image feature information are input into the sixth classifier, and a vehicle type recognition result is obtained by output, wherein the classification recognition labels of the sixth classifier include SUV, three-box sedan, and two-box sedan.

6. The occupancy identification device in the smart station is characterized by: The device is connected to a placeholder recognition system, the system includes a wake-up module, an image recognition module, a wireless recognition module and an alarm module, and the device includes: A determination and awakening unit, used to activate the awakening module and awaken the image recognition module and the wireless recognition module when the vehicle approaches the target parking space; A vehicle entry recognition unit, used to collect time-series images through the image recognition module, perform vehicle entry recognition, and obtain a vehicle entry recognition result; A vehicle identification unit, configured to, when the vehicle entry identification result is a vehicle entry, perform vehicle identification on the time-series image to obtain an image vehicle identification result, and identify a wireless tag of the vehicle entering through the wireless identification module to obtain a wireless vehicle identification result; An alarm determination unit is used to determine an occupancy recognition result according to the image vehicle recognition result and the wireless vehicle recognition result, and determine whether to start an alarm module to alarm or not according to the occupancy recognition result; Wherein, when the vehicle entry recognition result is vehicle entry, vehicle recognition is performed on the time-series image to obtain an image vehicle recognition result, and a wireless tag of the vehicle entering is recognized by the wireless recognition module to obtain a wireless vehicle recognition result, including: Extract the latest N images in the time series image, input them into the second deep learning model, and identify and obtain the license plate recognition result, the license plate presence recognition result and the vehicle model recognition result, wherein the second deep learning model includes a feature extraction network, a time series extraction network, a fourth classifier, a fifth classifier and a sixth classifier; When the license plate recognition result, the license plate in position recognition result and the vehicle type recognition result are a preset license plate, a license plate in position and a preset vehicle type, an image vehicle recognition result including the owner's vehicle is generated; otherwise, an image vehicle recognition result of a non-owner's vehicle is generated; Identify the wireless tag entering the vehicle through the wireless identification module, if there is a preset wireless tag, generate a wireless vehicle identification result including the vehicle owner's authorization, if not, generate a wireless vehicle identification result including the vehicle owner's unauthorized; The method of obtaining a placeholder recognition result by distinguishing the image vehicle recognition result and the wireless vehicle recognition result, and determining whether to start the alarm module to alarm or not according to the placeholder recognition result includes: When the image vehicle recognition result is the owner's vehicle and the wireless vehicle recognition result is the owner's authorization, it is determined that a non-occupied placeholder recognition result is obtained; When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is authorized by the owner, determining the placeholder recognition result of the placeholder authorized by the owner; When the image vehicle recognition result is a vehicle not owned by the owner and the wireless vehicle recognition result is an unauthorized vehicle by the owner, determining and obtaining an occupied place recognition result; When the place-occupancy identification result is place-occupancy, it is determined that the alarm module is started to alarm; otherwise, it is determined that the alarm module is not started.

7. An electronic device, characterized in that: include: Memory for storing computer software programs; A processor is used to read and execute the computer software program, thereby implementing the occupancy identification method in the smart station as described in any one of claims 1-5.

8. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores a computer software program, which, when executed by a processor, implements the method for occupancy recognition in a smart station as described in any one of claims 1 to 5.