Electric vehicle ladder entering identification method, electronic equipment and storage medium

Abstract

The invention discloses an electric vehicle elevator entering recognition method, electronic equipment and a storage medium. The method comprises the steps that video streams in an elevator are collected in real time; when the existence of the electric vehicle is identified from the video stream, acquiring a first image frame from the video stream; sending the first image frame to a server; receiving an identification result sent by the server, wherein the identification result is a judgment result of the first image frame called by the server by the cloud large model; and when the identification result is that the electric vehicle exists, giving an alarm and preventing the elevator from closing the door. Through the above mode, when it is identified that there is an electric bicycle in the video stream, the screenshot is sent to the server to call the cloud large model for re-identification, through double identification, the accuracy of electric bicycle identification is improved, the occurrence of false alarm and false alarm conditions is reduced, and the cloud large model is an artificial intelligence model with large-scale parameter quantities, so that the identification efficiency is improved. A more accurate recognition result can be provided, and the recognition accuracy is further improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent identification, and in particular to a method for identifying an electric vehicle entering an elevator and related devices. Background Art

[0002] Camera intelligent recognition technology is a method that combines computer vision and deep learning to automatically recognize, analyze and understand images. This technology has a wide range of applications in many fields, including security monitoring, smart driving, medical diagnosis, smart homes and industrial automation.

[0003] With the popularity of electric bicycles, the problem of community management of electric bicycles has become increasingly prominent, mainly in terms of charging safety of electric bicycles. In order to charge their electric bicycles conveniently, many electric bicycle owners will ride their electric bicycles directly back home to charge them.

[0004] The current camera intelligent recognition technology has low accuracy in identifying electric vehicles. Summary of the invention

[0005] The embodiments of the present application provide a method for identifying an electric vehicle entering an elevator, an electronic device, and a storage medium, which are used to improve the accuracy of identification.

[0006] In view of this, the present application provides, on one hand, a method for identifying an electric vehicle entering an elevator, comprising:

[0007] Real-time collection of video streams in elevators;

[0008] When an electric vehicle is identified in the video stream, a first image frame is acquired from the video stream;

[0009] Sending the first image frame to the server;

[0010] Receive the recognition result sent by the server, where the recognition result is the judgment result of the server calling the cloud-based large model on the first image frame;

[0011] When the recognition result is that there is an electric vehicle, an alarm is issued and the elevator door is prevented from closing.

[0012] Optionally, the first image frame includes a plurality of second image frames, and when the recognition result is that an electric vehicle is present, issuing an alarm and preventing the elevator from closing the door includes:

[0013] When the recognition result indicates that an electric vehicle exists in each second image frame, an alarm is issued and the elevator door is prevented from closing.

[0014] Optionally, after issuing an alarm and preventing the elevator door from closing, the method further includes:

[0015] receiving a false alarm instruction, the false alarm instruction indicating false alarm information;

[0016] Send notifications to management based on false positives;

[0017] generating false positive data based on the false positive information;

[0018] The false positive data is sent to the server so that the server can retrain the cloud-based large model based on the false positive data.

[0019] Optionally, sending the first image frame to the server includes:

[0020] An identification request is sent to a server, where the identification request includes a first image frame and a request identifier, and the false alarm data includes the request identifier.

[0021] Another aspect of the present application provides a method for identifying an electric vehicle entering an elevator, comprising:

[0022] receiving a first image frame sent by a terminal device, where the first image frame is obtained by the terminal device in a video stream in an elevator in which an electric vehicle is identified;

[0023] Calling the cloud-based large model to judge the first image frame and obtain a recognition result, where the recognition result indicates whether there is an electric vehicle in the first image frame;

[0024] Send the recognition result to the terminal device.

[0025] Optionally, the first image frame includes multiple second image frames, and the cloud-based large model is called to judge the first image frame, and the recognition result obtained includes:

[0026] Calling the cloud-based large model to judge each second image frame, and obtaining a judgment result corresponding to each second image frame;

[0027] When the judgment result corresponding to each second image frame is that an electric vehicle exists, it is determined that the recognition result is that an electric vehicle exists.

[0028] Optionally, after sending the recognition result to the terminal device, the method further includes:

[0029] Receiving false alarm data from terminal devices;

[0030] Retrain the large cloud model based on false positive data.

[0031] Optionally, receiving a first image frame sent by a terminal device includes:

[0032] Receiving an identification request sent by a terminal device, the identification request includes a first image frame and a request identifier, and the false alarm data includes the request identifier;

[0033] Retraining the large cloud model based on false positive data includes:

[0034] The cloud-based large model is retrained based on the false positive data and the first image frame corresponding to the request identifier.

[0035] Another aspect of the present application provides an electric vehicle elevator entry recognition device, comprising:

[0036] A collection unit, used to collect the video stream in the elevator in real time;

[0037] An acquisition unit, configured to acquire a first image frame from the video stream when an electric vehicle is identified from the video stream;

[0038] A sending unit, configured to send the first image frame to a server;

[0039] A receiving unit, used to receive a recognition result sent by the server, where the recognition result is a judgment result of the server calling a cloud-based large model on the first image frame;

[0040] The alarm unit is used to issue an alarm and prevent the elevator door from closing when the recognition result is that an electric vehicle exists.

[0041] Another aspect of the present application provides an electric vehicle elevator entry recognition device, comprising:

[0042] A receiving unit, configured to receive a first image frame sent by a terminal device, the first image frame being acquired by the terminal device in a video stream in an elevator in which an electric vehicle is identified;

[0043] A recognition unit, used for calling a large cloud model to judge the first image frame and obtain a recognition result, wherein the recognition result indicates whether there is an electric vehicle in the first image frame;

[0044] The sending unit is used to send the recognition result to the terminal device.

[0045] Another aspect of the present application provides an electronic device, comprising:

[0046] memories, transceivers, processors, and bus systems;

[0047] Wherein, the memory is used to store programs;

[0048] The processor is used to execute the program in the memory, including executing the above-mentioned methods;

[0049] The bus system is used to connect the memory and the processor so that the memory and the processor can communicate with each other.

[0050] Another aspect of the present application provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer is enabled to execute the above-mentioned methods.

[0051] Another aspect of the present application provides a computer program product or a computer program, the computer program product or the computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the electronic device executes the methods provided by the above aspects.

[0052] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0053] The embodiment of the present application collects the video stream in the elevator in real time; when an electric vehicle is identified in the video stream, a first image frame is obtained from the video stream; the first image frame is sent to the server; the recognition result sent by the server is received, and the recognition result is the judgment result of the server calling the cloud-based large model on the first image frame; when the recognition result is that there is an electric vehicle, an alarm is issued and the elevator door is prevented from closing. In the above manner, when an electric vehicle is identified in the video stream, a screenshot is taken to the server to call the cloud-based large model for re-recognition. The double recognition improves the accuracy of electric bicycle recognition and reduces the occurrence of false alarms and misreporting. The cloud-based large model is an artificial intelligence model with a large number of parameters, which can provide more accurate recognition results and further improve the recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A system architecture diagram in an embodiment of the present application;

[0055] Figure 2 A schematic diagram of a flow chart of a method for identifying an electric vehicle entering an elevator in an embodiment of the present application;

[0056] Figure 3 A schematic diagram of determining a recognition result in an embodiment of the present application;

[0057] Figure 4 A schematic diagram of an electric vehicle identification process in an embodiment of the present application;

[0058] Figure 5 This is a schematic diagram of the structure of an electric vehicle elevator entry recognition device in an embodiment of the present application;

[0059] Figure 6 This is a schematic diagram of the structure of another electric vehicle elevator entry recognition device in an embodiment of the present application;

[0060] Figure 7 It is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0061] The embodiments of the present application provide a method for identifying an electric vehicle entering an elevator, an electronic device, and a storage medium, which are used to improve the accuracy of identification.

[0062] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0063] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein, for example. In addition, the terms "including" and "corresponding to" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0064] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0065] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0066] In addition, in order to better illustrate the present application, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.

[0067] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0068] With the popularity of electric bicycles, the problem of community management of electric bicycles has become increasingly prominent. It is mainly reflected in the charging safety of electric vehicles. In order to facilitate charging of electric vehicles, many electric vehicle owners will ride their electric vehicles directly back home to charge. At present, most scenarios are to install electric vehicle recognition smart cameras at the entrance of the elevator, and use smart cameras to identify electric bicycles. However, due to the configuration of the edge device of the smart camera, generally a relatively single small artificial intelligence (AI) cloud model is embedded in the edge device, and there are certain misreporting in the recognition rate. For example, sometimes bicycles and strollers are recognized as electric bicycles, thereby issuing false alarms.

[0069] In response to the above technical problems, the embodiment of the present application provides a method for identifying electric vehicles entering an elevator, by real-time acquisition of the video stream in the elevator; when an electric vehicle is identified from the video stream, a first image frame is obtained from the video stream; the first image frame is sent to a server; the recognition result sent by the server is received, and the recognition result is the judgment result of the server calling the cloud-based large model on the first image frame; when the recognition result is that an electric vehicle exists, an alarm is issued and the elevator door is prevented from closing. In the above manner, a screenshot is taken from the video stream in which the electric vehicle is identified and sent to the server for identification. When the server receives the recognition result that the electric vehicle exists through the cloud-based large model, the elevator door is prevented from closing. That is, by double-identifying electric vehicles, the accuracy and error rate of electric bicycle elevator entry alarm recognition are improved.

[0070] System architecture and scenario description of the application of the embodiments of the present disclosure

[0071] Figure 1 1 is a system architecture diagram used in the electric vehicle elevator entry recognition method according to an embodiment of the present disclosure, which includes a terminal 140, an Internet 130, a gateway 120, a server 110, and the like.

[0072] The terminal 140 includes a variety of device forms with a display screen, such as a desktop computer, a laptop computer, a PDA (personal digital assistant), a mobile phone, a vehicle-mounted terminal, a home theater terminal, a dedicated terminal, an intelligent voice interaction device, an intelligent home appliance or an aircraft. In addition, it can be a single device or a collection of multiple devices. The terminal 140 can communicate with the Internet 130 in a wired or wireless manner to exchange data. The terminal of the embodiment of the present application can be a smart camera, which can be a device that includes both the image acquisition and image analysis functions, or a combination of a camera for acquiring images and a processing device for analyzing images, which is not limited here.

[0073] The server 110 refers to a computer system that can provide certain services to the terminal 140. Compared with the ordinary terminal 140, the server 110 has higher requirements in terms of stability, security, performance, etc. The server 110 can be a high-performance computer in a network platform, a cluster of multiple high-performance computers, a part of a high-performance computer (such as a virtual machine), a combination of parts of multiple high-performance computers (such as virtual machines), etc.

[0074] The gateway 120 is also called an internetwork connector or a protocol converter. The gateway realizes network interconnection at the transport layer and is a computer system or device that acts as a converter. The gateway is a translator between two systems that use different communication protocols, data formats or languages, or even completely different architectures. At the same time, the gateway can also provide filtering and security functions. The message sent by the terminal 140 to the server 110 must be sent to the corresponding server 110 through the gateway 120. The message sent by the server 110 to the terminal 140 must also be sent to the corresponding terminal 140 through the gateway 120.

[0075] In the method for identifying electric vehicles entering an elevator provided by the embodiment of the present disclosure, the terminal 140 collects the video stream in the elevator in real time; when the presence of an electric vehicle is identified from the video stream, the terminal 140 obtains the first image frame from the video stream; the terminal 140 sends the first image frame to the server 110; the server 110 calls the cloud-based large model to judge the first image frame and obtains a recognition result; the server 110 sends the recognition result to the terminal 140; when the recognition result is that there is an electric vehicle, the terminal 140 issues an alarm and prevents the elevator door from closing.

[0076] The following describes a method for identifying electric vehicles entering an elevator provided in an embodiment of the present application. Taking a terminal as an intelligent camera as an example, the method is as follows:

[0077] See also Figure 2 ,like Figure 2 The figure is a flow chart of a method for identifying an electric vehicle entering an elevator provided in an embodiment of the present application, including:

[0078] Step 201. The smart camera collects the video stream in the elevator in real time.

[0079] In this embodiment, the smart camera can be installed inside the elevator to continuously monitor the situation inside the elevator. The video stream is a continuous sequence of images, that is, the smart camera starts working and captures the dynamic images inside the elevator in real time to form a continuous video stream. The smart camera can monitor the activities in the elevator by capturing the video stream in real time.

[0080] Step 202: When an electric vehicle is identified in the video stream, the smart camera acquires a first image frame from the video stream.

[0081] In this embodiment, the smart camera may have a built-in image recognition algorithm, which can continuously analyze the video stream in real time and use the built-in image recognition algorithm to detect specific objects, such as electric vehicles. Once the algorithm detects the presence of an electric vehicle, the camera will capture the current image frame from the video stream as the first image frame, which contains the image information of the electric vehicle. The first image frame obtained may also be accompanied by a timestamp to record the specific time when the electric vehicle was identified.

[0082] The way to identify video streams through image recognition algorithms can be through deep learning algorithms, such as convolutional neural networks (CNN) or target detection models such as YOLO (You Only Look Once), where CNN can automatically learn feature representations from raw image data, capture spatial local features in images, and maintain translation invariance. The YOLO algorithm can be trained directly from image pixels to bounding boxes and category probabilities without complex preprocessing or feature extraction steps, and can accurately find the location of objects in a given picture and mark the object category. Among them, identifying electric vehicles can be based on features such as shape, size, and color of electric vehicles.

[0083] Step 203: The smart camera sends the first image frame to the server.

[0084] In this embodiment, after capturing the first image frame containing the electric vehicle, the smart camera can send the image frame to the server through the network. This transmission process may be completed through a wireless network or a wired network, depending on the configuration of the smart camera and the network.

[0085] Step 204. The server calls the cloud-based large model to judge the first image frame and obtain a recognition result, which indicates whether there is an electric vehicle in the first image frame.

[0086] In this embodiment, after receiving the first image frame, the server uses the pre-trained cloud-based large model to analyze the first image frame. The cloud-based large model can extract features from the first image frame through its trained network structure, and analyze the extracted features to identify whether there is an electric vehicle in the first image frame. When the cloud-based large model does not identify the electric vehicle in the first image frame, the recognition result is that there is no electric vehicle in the image, and when the cloud-based large model identifies the electric vehicle in the first image frame, the recognition result is that there is an electric vehicle in the image.

[0087] Cloud-based big models are AI models with large-scale parameters deployed on cloud servers. They can process and analyze large amounts of data to provide more accurate recognition and analysis services. Cloud-based big models have powerful computing capabilities and complex algorithm structures, and can perform complex tasks such as image recognition and video analysis.

[0088] In one implementation, the first image frame includes multiple second image frames, and the cloud-based large model is called to judge the first image frame, and the recognition result obtained includes:

[0089] Calling the cloud-based large model to judge each second image frame, and obtaining a judgment result corresponding to each second image frame;

[0090] When the judgment result corresponding to each second image frame is that an electric vehicle exists, it is determined that the recognition result is that an electric vehicle exists.

[0091] In this implementation, the multiple second image frames may be multiple images captured by the smart camera from the video stream, that is, each image is recognized by the smart camera as an electric vehicle. The server can call the cloud-based large model to extract features from each second image frame, and analyze the extracted features to obtain a judgment result corresponding to each second image frame, that is, to determine whether an electric vehicle exists. Among them, since the judgment results of the cloud-based large model for each second image frame may be different, the embodiment of the present application can generate a recognition result indicating the existence of an electric vehicle only when the judgment result corresponding to each second image frame in the multiple second image frames is that an electric vehicle exists, that is, when the judgment result corresponding to any second image frame in the multiple second image frames is that there is no electric vehicle, a recognition result indicating that there is no electric vehicle is generated. Through multiple image recognition, the accuracy of cloud-based large model recognition is further improved.

[0092] Exemplarily, the cloud-based large model receives two images I MG1 and I MG2 uploaded by the AI ​​camera, identifies the two images respectively, performs AND operation on the identified image results, and obtains the judgment result I MG1(R) corresponding to I MG1 and the judgment result I MG2(R) corresponding to I MG2. The calculation formula of the recognition result is:

[0093] Result=I MG1(R)&&I MG2(R)

[0094] See also Figure 3As shown in the schematic diagram of determining the recognition result, picture 1 and picture 2 are respectively input into the cloud-based large model for judgment, and the judgment results corresponding to picture 1 and picture 2 are respectively saved, and the recognition result is determined based on the judgment result 1 corresponding to picture 1 and the judgment result 2 corresponding to picture 2. Only when the recognition of the two pictures returns that there is an electric bicycle, the recognition result is returned as T, indicating that there is an electric bicycle, otherwise the recognition result is returned as F, indicating that there is no electric bicycle.

[0095] It is described above that the recognition result sent by the server to the smart camera directly indicates whether there is an electric vehicle. In another embodiment, the server can also send the judgment result corresponding to each second image frame to the smart camera. The smart camera determines that there is an electric vehicle only when it determines that the judgment result corresponding to each second image frame is that there is an electric vehicle.

[0096] Step 205: The server sends the recognition result to the smart camera.

[0097] In this embodiment, after the server processes the first image frame and obtains the recognition result, the recognition result can be sent back to the smart camera through the network. The recognition result can be a message describing whether there is an electric vehicle, or an image frame including a bounding box, confidence and category label of the electric vehicle, or a simple signal indicating whether there is an electric vehicle in the image.

[0098] Step 206. When the recognition result shows that there is an electric vehicle, the smart camera issues an alarm and prevents the elevator door from closing.

[0099] In this embodiment, after receiving the recognition result from the server, the smart camera can determine whether the cloud-based large model recognizes an electric vehicle. If the recognition result indicates that there is an electric vehicle in the first image frame, the camera can trigger an alarm mechanism, such as issuing a sound or light warning. At the same time, the smart camera can send a signal to the elevator control system to prevent the elevator door from closing until the electric vehicle is detected to be removed from the elevator. The smart camera can record all relevant judgment results and operations, including alarms and elevator control events, and send alarm information and event details to the security monitoring center or elevator maintenance personnel.

[0100] In a possible implementation manner, after issuing an alarm and preventing the elevator door from closing, the method further includes:

[0101] The smart camera receives a false alarm instruction, and the false alarm instruction indicates false alarm information;

[0102] Smart cameras send notifications to managers based on false alarms;

[0103] Smart cameras generate false alarm data based on false alarm information;

[0104] The smart camera sends false alarm data to the server;

[0105] The server retrains the large cloud model based on the false positive data.

[0106] In this implementation, the false alarm instruction may be that when the smart camera mistakenly identifies that there is no electric vehicle and issues an alarm, the smart camera receives an instruction issued by an elevator passenger, and the instruction may indicate the time and location of the false alarm, that is, the false alarm information includes the false alarm time and the false alarm location. Exemplarily, a false alarm button may be configured at the entrance of the elevator, and the passenger generates the false alarm instruction by triggering the false alarm button.

[0107] The smart camera can generate a notification based on the false alarm information. The notification can be in the form of a false alarm prompt alarm, or it can send status information to the management personnel through a preset communication channel (such as email, text message, APP push, etc.), such as the time of the error, the location of the false alarm and the possible cause. Among them. The false alarm instruction can also be triggered by the management personnel when viewing the first image frame corresponding to the recognition result. That is, when the management personnel determines that the first image frame corresponding to the recognition result indicating the existence of an electric vehicle does not include an electric vehicle, a false alarm instruction can be sent to the smart camera through a specific interface or console. At this time, the smart camera does not need to send a notification to the management personnel.

[0108] Smart cameras can record detailed information of false alarm events, including false alarm time, environmental conditions (such as light, weather), etc., to generate false alarm data.

[0109] These false positive data are then packaged and transmitted to the central server via the network, and the server retrains the cloud-based large model based on the false positive data. This process may include adjusting the model's parameters, increasing the proportion of negative samples, optimizing the loss function, etc., to reduce the occurrence of false positives in the future.

[0110] During the retraining process, data augmentation technology can be used to transform false positive images and generate more negative samples to enhance the model's ability to identify false positives.

[0111] In this way, large models in the cloud can continuously learn and adapt, improving their accuracy and robustness and reducing false positives.

[0112] In a possible implementation manner, receiving a first image frame sent by a terminal device includes:

[0113] Receiving an identification request sent by a terminal device, the identification request includes a first image frame and a request identifier, and the false alarm data includes the request identifier;

[0114] Retraining the large cloud model based on false positive data includes:

[0115] The cloud-based large model is retrained based on the false positive data and the first image frame corresponding to the request identifier.

[0116] In this implementation, the identification request may be a request initiated by the smart camera to the server to identify the first image frame. The identification request may include a request identifier, and the request identifier uniquely corresponds to the first image frame.

[0117] When the smart camera determines a false alarm image frame, it can generate false alarm data based on the request identifier corresponding to the image frame. That is, when the server receives the false alarm data, it can determine the first image frame that has been identified based on the request identifier in the false alarm data, and then the first image frame can be used as a negative sample to retrain the large cloud model.

[0118] Among them, for negative samples, you can copy and scale them to the background. You can also search for some similar images to add to the training set, and then retrain the large cloud model with the training set to improve the model's resolution ability.

[0119] The present application embodiment is described in conjunction with a specific implementation scenario. Figure 4The schematic diagram of the electric vehicle identification process shown in the figure includes step 401. The smart camera obtains the video stream of the elevator entrance in real time, that is, continuously shoots the elevator entrance. Step 402. The smart camera identifies whether there is an electric vehicle in the video stream based on the recognition algorithm. Step 403. When there is no electric vehicle in the video stream, the recognition process ends, and the smart camera continues to identify the next video stream obtained in real time. Step 404. When there is an electric vehicle in the video stream, two pictures can be screenshot respectively, picture 1 and picture 2, for example, the video stream of the electric vehicle is screenshotted 2 seconds apart. Step 405. The smart camera can encapsulate the two pictures into request data, and the request data can include a request identifier, a camera identifier, a time, picture 1 and picture 2. The camera identifier uniquely indicates the smart camera, that is, the server can support the recognition requests sent by multiple cameras, and the data of each camera is saved separately based on the camera identifier. Step 406. The smart camera can call the cloud-based large model in the server to identify picture 1 and picture 2 respectively, and obtain the judgment result corresponding to picture 1 and the judgment result corresponding to picture 2. Provide a picture upload service interface on the cloud big model service, supporting the upload of picture 1 and picture 2. Step 407. The server generates a recognition result based on the judgment result corresponding to picture 1 and the judgment result corresponding to picture 2, and sends it to the smart camera. The recognition result indicates the existence of an electric vehicle only when the judgment result corresponding to picture 1 and the judgment result corresponding to picture 2 are both that an electric vehicle exists. At the same time, the recognition result also includes a request identifier, a camera identifier, a time, picture 1 and picture 2, that is, the recognition result is sent to the corresponding smart camera through the camera identifier, and the smart camera determines the corresponding recognition request based on the request identifier and time. Step 408. The smart camera determines whether the recognition result indicates the existence of an electric vehicle. When the recognition result indicates that there is no electric vehicle, execute step 403, the recognition process ends, and the smart camera continues to recognize the next video stream obtained in real time. Step 409. Issue an alarm and prevent the elevator from closing the door. Step 410. The user presses the false alarm button when determining a false alarm. Step 411. The smart camera notifies the manager of a false alarm, and the smart camera sends the false alarm information to the monitoring center or the manager's device through the built-in communication module. Step 412. Generate false alarm data and send the false alarm data to the server. The false alarm data may include the camera ID and the request ID. Step 413. The cloud-based large model saves the false alarm data. Step 414. Manually review whether the image corresponding to the false alarm data is true. Step 415. If it is a true false alarm, retrain the cloud-based large model based on the image corresponding to the false alarm data.

[0120] The embodiment of the present application collects the video stream in the elevator in real time; when an electric vehicle is identified in the video stream, a first image frame is obtained from the video stream; the first image frame is sent to the server; the recognition result sent by the server is received, and the recognition result is the judgment result of the server calling the cloud-based large model on the first image frame; when the recognition result is that there is an electric vehicle, an alarm is issued and the elevator door is prevented from closing. In the above manner, when an electric vehicle is identified in the video stream, a screenshot is taken to the server to call the cloud-based large model for re-recognition. The double recognition improves the accuracy of electric bicycle recognition and reduces the occurrence of false alarms and misreporting. The cloud-based large model is an artificial intelligence model with a large number of parameters, which can provide more accurate recognition results and further improve the recognition accuracy.

[0121] The above describes the method for identifying an electric vehicle entering an elevator. The following describes a device for executing the method.

[0122] See also Figure 5 The schematic diagram of the structure of a device for identifying electric vehicle entering an elevator is shown, and the device 50 includes:

[0123] The acquisition unit 501 is used to acquire the video stream in the elevator in real time;

[0124] An acquisition unit 502 is configured to acquire a first image frame from the video stream when an electric vehicle is identified in the video stream;

[0125] A sending unit 503, configured to send the first image frame to a server;

[0126] A receiving unit 504 is used to receive a recognition result sent by the server, where the recognition result is a judgment result of the server calling a cloud-based large model on the first image frame;

[0127] The alarm unit 505 is used to issue an alarm and prevent the elevator from closing the door when the recognition result shows that there is an electric vehicle.

[0128] Optionally, the first image frame includes a plurality of second image frames, and the alarm unit 505 is specifically used for:

[0129] When the recognition result indicates that an electric vehicle exists in each second image frame, an alarm is issued and the elevator door is prevented from closing.

[0130] Optionally, the receiving unit 504 is further configured to:

[0131] receiving a false alarm instruction, the false alarm instruction indicating false alarm information;

[0132] The sending unit 503 is further used for:

[0133] Send notifications to management based on false positives;

[0134] The apparatus 50 further includes a generating unit 506, which is specifically configured to:

[0135] generating false positive data based on the false positive information;

[0136] The sending unit 503 is further used for:

[0137] The false positive data is sent to the server so that the server can retrain the cloud-based large model based on the false positive data.

[0138] Optionally, the receiving unit 504 is specifically configured to:

[0139] An identification request is sent to a server, where the identification request includes a first image frame and a request identifier, and the false alarm data includes the request identifier.

[0140] See also Figure 6 ,like Figure 6 FIG. 6 is a schematic diagram of another electric vehicle elevator entry recognition device provided in an embodiment of the present application. The device 60 includes:

[0141] A receiving unit 601 is used to receive a first image frame sent by a terminal device, where the first image frame is obtained by the terminal device in a video stream in an elevator in which an electric vehicle is identified;

[0142] The recognition unit 602 is used to call the cloud-based large model to judge the first image frame and obtain a recognition result, where the recognition result indicates whether there is an electric vehicle in the first image frame;

[0143] The sending unit 603 is used to send the recognition result to the terminal device.

[0144] Optionally, the first image frame includes a plurality of second image frames, and the identification unit 602 is specifically configured to:

[0145] Calling the cloud-based large model to judge each second image frame, and obtaining a judgment result corresponding to each second image frame;

[0146] When the judgment result corresponding to each second image frame is that an electric vehicle exists, it is determined that the recognition result is that an electric vehicle exists.

[0147] Optionally, the receiving unit is further configured to:

[0148] Receiving false alarm data from terminal devices;

[0149] The apparatus 60 further includes a training unit 604, which is specifically configured to:

[0150] Retrain the large cloud model based on false positive data.

[0151] Optionally, the receiving unit 601 is specifically configured to:

[0152] Receiving an identification request sent by a terminal device, the identification request includes a first image frame and a request identifier, and the false alarm data includes the request identifier;

[0153] The training unit 604 is specifically used for:

[0154] The cloud-based large model is retrained based on the false positive data and the first image frame corresponding to the request identifier.

[0155] Figure 7 : This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device 300 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 322 (for example, one or more processors) and a memory 332, and one or more storage media 330 (for example, one or more mass storage devices) storing application programs 342 or data 344. Among them, the memory 332 and the storage medium 330 can be short-term storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the electronic device. Furthermore, the central processing unit 322 can be configured to communicate with the storage medium 330 to execute a series of instruction operations in the storage medium 330 on the electronic device 300.

[0156] The electronic device 300 may also include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input and output interfaces 358, and / or one or more operating systems 341, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0157] The steps performed by the terminal device or the server in the above embodiment can be based on the Figure 7 The electronic device structure shown.

[0158] A computer-readable storage medium is also provided in an embodiment of the present application, on which a computer program is stored. When the computer program is executed by a processor, the steps of the methods described in the above embodiments are implemented.

[0159] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0160] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0161] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0162] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0163] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable an electronic device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0164] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for identifying an electric vehicle entering an elevator, characterized in that: include: Real-time collection of video streams in elevators; When the presence of an electric vehicle is identified from the video stream, acquiring a first image frame from the video stream; Sending the first image frame to a server; Receiving a recognition result sent by the server, where the recognition result is a judgment result of the server calling a cloud-based large model on the first image frame; When the identification result is that an electric vehicle exists, an alarm is issued and the elevator door is prevented from closing.

2. The method according to claim 1, characterized in that: The first image frame includes a plurality of second image frames, and when the recognition result is that an electric vehicle is present, issuing an alarm and preventing the elevator from closing the door includes: When the recognition result indicates that an electric vehicle exists in each second image frame, an alarm is issued and the elevator door is prevented from closing.

3. The method according to claim 1, characterized in that: After issuing an alarm and preventing the elevator door from closing, the method further includes: receiving a false alarm instruction, wherein the false alarm instruction indicates false alarm information; sending a notification to an administrator based on the false positive information; generating false alarm data according to the false alarm information; The false alarm data is sent to the server, so that the server retrains the cloud-based large model according to the false alarm data.

4. A method for identifying an electric vehicle entering an elevator, characterized in that: include: receiving a first image frame sent by a terminal device, where the first image frame is obtained by the terminal device in a video stream in an elevator in which an electric vehicle is identified; Calling a large cloud model to judge the first image frame to obtain a recognition result, wherein the recognition result indicates whether there is an electric vehicle in the first image frame; The identification result is sent to the terminal device.

5. The method according to claim 4, characterized in that The first image frame includes a plurality of second image frames, and the calling of the cloud-based large model to judge the first image frame to obtain a recognition result includes: The calling of the cloud-based large model to judge each second image frame to obtain a judgment result corresponding to each second image frame; When the judgment result corresponding to each of the second image frames is that an electric vehicle exists, it is determined that the recognition result is that an electric vehicle exists.

6. The method according to claim 4, characterized in that After sending the recognition result to the terminal device, the method further includes: receiving false alarm data from the terminal device; The cloud-based large model is retrained according to the false alarm data.

7. The method according to claim 6, characterized in that The receiving terminal device sends a first image frame including: receiving an identification request sent by the terminal device, wherein the identification request includes a first image frame and a request identifier, and the false alarm data includes the request identifier; The retraining of the cloud-based large model according to the false alarm data includes: The cloud-based large model is retrained according to the false alarm data and the first image frame corresponding to the request identifier.

8. An electronic device, characterized in that: include: memories, transceivers, processors, and bus systems; Wherein, the memory is used to store programs; The processor is used to execute the program in the memory, including executing the method according to any one of claims 1 to 7; The bus system is used to connect the memory and the processor so that the memory and the processor can communicate with each other.

9. A computer-readable storage medium comprising instructions, which, when executed on a computer, causes the computer to perform the method according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, wherein the computer program is read and executed by a processor of an electronic device, so that the electronic device executes the method according to any one of claims 1 to 7.

Images