Article tracking method and device based on intelligent security camera

Abstract

The invention belongs to the technical field of data processing, and particularly relates to an article tracking method and device based on an intelligent security camera. The method comprises the following steps: S1, determining an article to be tracked; s2, matching a video frame set containing an article needing to be tracked from video frames pre-stored in a database; s3, in the video frame set, according to the time sequence of the video frames, determining the motion trail of an object needing to be tracked; s4, according to the motion trail, determining the position of the article which needs to be tracked and is finally shot or the position of the article which needs to be tracked and is indicated in the video frame outside the shooting range, and taking the position as a positioning result of the article which needs to be tracked; and S5, sending the positioning result and the motion trail to the user. The method can help the user to quickly find the forgotten article, saves time and energy, improves the life efficiency, and improves the user experience.

Description

Technical Field

[0001] The present application belongs to the field of data processing technology, and in particular relates to an object tracking method and device based on an intelligent security camera. Background Art

[0002] In a home environment, users often forget the specific location of items due to the fast pace of life and high pressure, resulting in a lot of time and energy being spent looking for these items.

[0003] In the existing technology, object tracking and positioning technology mainly relies on RFID (radio frequency identification), Bluetooth or GPS and other technologies. Although these technologies can achieve object tracking and positioning, they still have some shortcomings in practical applications: RFID tags need to be attached to items in advance, and their recognition range is limited. They are mainly suitable for tracking items at close range and in a small area. In addition, the cost of RFID tags is high, which is not conducive to large-scale promotion. Although Bluetooth trackers can achieve remote tracking of items, they consume a lot of power and require regular battery replacement. In addition, the recognition accuracy and stability are affected by environmental factors. GPS trackers are mainly used for tracking and locating outdoor items. They are expensive and not suitable for indoor environments. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides an object tracking method and device based on intelligent security cameras, which realizes automatic identification, tracking and positioning of objects in the home by combining the real-time monitoring capability of home security cameras, the big data storage and processing capability of cloud databases, the precise identification capability of AI visual recognition algorithms and the data mining capability of big data analysis technology.

[0005] The first aspect of the present application provides an object tracking method based on an intelligent security camera, which mainly includes: Step S1, determining the items to be tracked; Step S2: Matching a set of video frames containing objects to be tracked from video frames pre-stored in a database; Step S3, in the video frame set, determining the motion trajectory of the object to be tracked according to the time sequence of the video frames; Step S4, determining the last captured position of the object to be tracked or the position of the object indicated in the video frame where the object to be tracked is taken out of the camera range according to the motion trajectory, as a positioning result of the object to be tracked; Step S5: Send the positioning result and motion trajectory to the user.

[0006] Preferably, step S1 further comprises: Step S11: Determine the items to be tracked according to the description of the items to be tracked input by the user.

[0007] Preferably, in step S11, the description of the item to be tracked input by the user includes a text description or a picture description.

[0008] Preferably, step S1 further comprises: Step S12: determine the items that need to be tracked based on the statistical results and the current time. The statistical results refer to the number of times a user uses a specific item within a specific time period. When the number exceeds a set value and the current time reaches the start time of the specific time period, the specific item is used as an item that needs to be tracked.

[0009] Preferably, step S5 further comprises: The location result, time and situation of the item to be tracked are combined to generate a text description.

[0010] The second aspect of the present application provides an object tracking device based on an intelligent security camera, which mainly includes: An item determination module, used to determine items that need to be tracked; The object video frame query module is used to match the video frame set containing the object to be tracked from the video frames pre-stored in the database; An object motion trajectory generating module, used to determine the motion trajectory of the object to be tracked in the video frame set according to the time sequence of the video frames; An object position determination module is used to determine the last captured position of the object to be tracked or the position of the object indicated in the video frame where the object to be tracked is taken out of the camera range according to the motion trajectory, as a positioning result of the object to be tracked; The feedback module is used to send the positioning result and motion trajectory to the user.

[0011] Preferably, the item determination module includes: The description-based item determination unit is used to determine the items to be tracked according to the description of the items to be tracked input by the user.

[0012] Preferably, the description of the item to be tracked input by the user includes a text description or a picture description.

[0013] Preferably, the item determination module includes: The item determination unit based on statistical data is used to determine the items that need to be tracked according to the statistical results and the current time. The statistical results refer to the number of times a user uses a specific item within a specific time period. When the number exceeds a set value and the current time reaches the start time of the specific time period, the specific item is regarded as an item that needs to be tracked.

[0014] Preferably, the feedback module includes: The text organization unit is used to combine the location result, time and situation of the item to be tracked into a text description.

[0015] This application can help users quickly find forgotten items, save time and energy, improve life efficiency, and enhance user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of a preferred embodiment of the object tracking method based on the intelligent security camera of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the implementation of this application clearer, the technical scheme in the implementation of this application will be described in more detail in combination with the drawings in the implementation of this application. In the drawings, the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The described implementation is a part of the implementation of this application, not all of the implementations. The implementation described below with reference to the drawings is exemplary and is intended to be used to explain this application, and cannot be understood as a limitation on this application. Based on the implementation in this application, all other implementations obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. The implementation of this application is described in detail below in combination with the drawings.

[0018] The first aspect of the present application provides an object tracking method based on an intelligent security camera, such as Figure 1 As shown, it mainly includes: Step S1, determining the items to be tracked; Step S2: Matching a set of video frames containing objects to be tracked from video frames pre-stored in a database; Step S3, in the video frame set, determining the motion trajectory of the object to be tracked according to the time sequence of the video frames; Step S4, determining the last captured position of the object to be tracked or the position of the object indicated in the video frame where the object to be tracked is taken out of the camera range according to the motion trajectory, as a positioning result of the object to be tracked; Step S5: Send the positioning result and motion trajectory to the user.

[0019] In order to achieve the above steps, this application configures home security cameras, cloud servers and user terminals in hardware. Among them, the home security camera has a high-definition camera, image processing module and network communication function, which is used to capture the video images in the home in real time and transmit them to the cloud; the cloud server includes a cloud database, AI visual recognition engine, big data analysis module, etc., which is used to store video data, process image recognition tasks, analyze user behavior patterns and object movement trajectories; the user terminal (such as a smartphone APP) is used to provide a user interface, allowing users to input search requests, receive search results and view object location information.

[0020] It should also be noted that the home security camera of the present application continuously captures the video images in the home and uploads the video data to the cloud server in real time. The cloud server compresses and encodes the received video data and stores it in the cloud database for subsequent analysis.

[0021] The present application first determines the items to be tracked in step S1. In some optional implementations, step S1 further includes: step S11, determining the items to be tracked according to the description of the items to be tracked input by the user. In an alternative implementation, step S1 further includes: step S12, determining the items to be tracked according to the statistical results and the current time, the statistical results refer to the number of times the user uses a specific item in a specific time period. When the number exceeds a set value and the current time reaches the start time of the specific time period, the specific item is used as an item to be tracked.

[0022] Two methods are given above to determine the items that need to be tracked. One is to actively accept user query instructions and determine the items according to the user instructions. The other is to automatically predict the items that the user may query based on statistical results.

[0023] Regarding the method of actively accepting user query instructions, in some optional implementations, the description of the item to be tracked input by the user includes a text description or a picture description.

[0024] In this embodiment, the user inputs the information of the item to be tracked (such as pictures, text descriptions or voice commands) through the user terminal and sends it to the cloud server. For text descriptions, such as "I want to find my keys" or "Help me find my keys", the system will automatically extract the keyword "keys" as the item to be tracked; for picture descriptions, such as uploading a photo of the key, the system can use deep learning models (such as YOLO, SSD) to detect and classify various objects in the video image and identify the key as the item to be tracked.

[0025] For automatic prediction based on statistical results, the system can learn the user's daily behavior habits, predict and mark items that may be forgotten in advance, and improve tracking efficiency. For example, if the statistical results show that the user uses the car key at 7 o'clock every morning, then before 7 o'clock, the system will automatically treat the car key as an item that needs to be tracked and perform subsequent query steps.

[0026] In some optional implementations, the statistical results or statistical methods are modified based on the query effectiveness of the items to be tracked as fed back by users, or the item tracking parameters are modified based on the query accuracy of the items to be tracked as fed back by users.

[0027] In this embodiment, a user feedback mechanism is introduced. One is to modify the accuracy of the automatic prediction of the items to be tracked based on the statistical results. Assuming that the items to be tracked are car keys, and the user feedback shows that they don’t really care where the car keys are, or the user feedback shows that they only need to use the car keys during work, these two feedback results will affect the statistical results and statistical methods respectively. For the other user feedback mechanism, it is essentially a correction to the model used in the subsequent steps S3-S4, that is, to continuously optimize the performance of the AI ​​visual recognition engine and the big data analysis model, and improve the recognition accuracy and tracking speed.

[0028] After the object is determined, step S2 is used to query the relevant video frames in the cloud server. The cloud server uses the AI ​​visual recognition engine to scan the video data stored in the cloud database, identify the objects that match the user input information, and mark them. For example, each video frame that identifies the object to be tracked is recorded as a record, and each record includes timestamp, object type, location coordinates and other information. The identified object information and its location information are stored in the cloud database and kept updated in real time. It should be noted that when querying each video frame, the accuracy of target detection can be improved by denoising, enhancement and other methods.

[0029] In step S3, the cloud server uses a big data analysis module to analyze the movement trajectory of the marked object, and in step S4, further determines its last appearance location in the home or whether it has been taken out of the home.

[0030] Based on the video frames, detailed information such as the last location and time of the object can be determined. Combined with the visual recognition algorithm, the cloud server can further refine the object location information, such as the specific room, corner or next to the furniture.

[0031] Finally, in step S5, the cloud server sends the tracking and positioning results to the user terminal, and displays the item location information, movement trajectory, and timestamp in a graphical interface. Users can quickly find the forgotten items based on the feedback information, or take further actions (such as going out to look for them). In this step, the retrieved information will be displayed on the terminal's APP interface, including detailed information such as the location where the item last appeared (for example, next to the sofa in the living room), time, and whether it was taken out of the room. At the same time, an indoor map can be generated and the location of the item can be marked for reference.

[0032] In some optional implementations, step S5 further includes: The location result, time and situation of the item to be tracked are combined to generate a text description.

[0033] In this embodiment, the situation mainly refers to the item described in step S4 being within the monitoring range or being taken out of the monitoring range. For example, when the situation is not within the monitoring range, the organized text description can be: Your car keys were last seen at 8:30 last night on the coffee table in the living room.

[0034] In an alternative implementation, the generated text description may also include prediction results based on statistics. For example, based on historical data analysis, it can roughly predict whether an item has been taken out of the room, thereby providing more detailed information. For example, the organized text description may be: Your car keys were last seen at 8:30 p.m. yesterday on the coffee table in the living room. According to statistical results, the keys are usually placed in the drawer of the coffee table.

[0035] This application uses cloud database and big data analysis technology to achieve intelligent management of household items, including item identification, tracking, positioning, and personalized service recommendations, thereby improving users' quality of life and happiness.

[0036] The second aspect of the present application provides an object tracking device based on an intelligent security camera corresponding to the above method, mainly comprising: An item determination module, used to determine items that need to be tracked; The object video frame query module is used to match the video frame set containing the object to be tracked from the video frames pre-stored in the database; An object motion trajectory generating module, used to determine the motion trajectory of the object to be tracked in the video frame set according to the time sequence of the video frames; An object position determination module is used to determine the last captured position of the object to be tracked or the position of the object indicated in the video frame where the object to be tracked is taken out of the camera range according to the motion trajectory, as a positioning result of the object to be tracked; The feedback module is used to send the positioning result and motion trajectory to the user.

[0037] In some optional implementations, the item determination module includes: The description-based item determination unit is used to determine the items to be tracked according to the description of the items to be tracked input by the user.

[0038] In some optional implementations, the description of the item to be tracked input by the user includes a text description or a picture description.

[0039] In some optional implementations, the item determination module includes: The item determination unit based on statistical data is used to determine the items that need to be tracked according to the statistical results and the current time. The statistical results refer to the number of times a user uses a specific item within a specific time period. When the number exceeds a set value and the current time reaches the start time of the specific time period, the specific item is regarded as an item that needs to be tracked.

[0040] In some optional implementations, the feedback module includes: The text organization unit is used to combine the location result, time and situation of the item to be tracked into a text description.

[0041] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. An object tracking method based on an intelligent security camera, characterized in that: include: Step S1, determining the items to be tracked; Step S2: matching a set of video frames containing objects to be tracked from video frames pre-stored in a database; Step S3, in the video frame set, determining the motion trajectory of the object to be tracked according to the time sequence of the video frames; Step S4, determining the last captured position of the object to be tracked or the position of the object indicated in the video frame where the object to be tracked is taken out of the camera range according to the motion trajectory, as a positioning result of the object to be tracked; Step S5: Send the positioning result and motion trajectory to the user.

2. The object tracking method based on the intelligent security camera according to claim 1, characterized in that: Step S1 further comprises: Step S11: Determine the items to be tracked according to the description of the items to be tracked input by the user.

3. The object tracking method based on the intelligent security camera as claimed in claim 2, characterized in that: In step S11, the description of the item to be tracked input by the user includes a text description or a picture description.

4. The object tracking method based on the intelligent security camera as claimed in claim 1, characterized in that: Step S1 further comprises: Step S12: determine the items that need to be tracked based on the statistical results and the current time. The statistical results refer to the number of times a user uses a specific item within a specific time period. When the number exceeds a set value and the current time reaches the start time of the specific time period, the specific item is used as an item that needs to be tracked.

5. The object tracking method based on the intelligent security camera as claimed in claim 1, characterized in that: Step S5 further comprises: The location result, time and situation of the item to be tracked are combined to generate a text description.

6. An object tracking device based on an intelligent security camera, characterized in that: include: An item determination module, used to determine items that need to be tracked; The object video frame query module is used to match the video frame set containing the object to be tracked from the video frames pre-stored in the database; An object motion trajectory generating module, used to determine the motion trajectory of the object to be tracked in the video frame set according to the time sequence of the video frames; An object position determination module is used to determine the last captured position of the object to be tracked or the position of the object indicated in the video frame where the object to be tracked is taken out of the camera range according to the motion trajectory, as a positioning result of the object to be tracked; The feedback module is used to send the positioning result and motion trajectory to the user.

7. The object tracking device based on the intelligent security camera as claimed in claim 6, characterized in that: The item determination module comprises: The description-based item determination unit is used to determine the items to be tracked according to the description of the items to be tracked input by the user.

8. The object tracking device based on the intelligent security camera as claimed in claim 7, characterized in that: The description of the item to be tracked input by the user includes a text description or a picture description.

9. The object tracking device based on the intelligent security camera as claimed in claim 6, characterized in that: The item determination module comprises: The item determination unit based on statistical data is used to determine the items that need to be tracked according to the statistical results and the current time. The statistical results refer to the number of times a user uses a specific item within a specific time period. When the number exceeds a set value and the current time reaches the start time of the specific time period, the specific item is regarded as an item that needs to be tracked.

10. The object tracking device based on the intelligent security camera as claimed in claim 6, characterized in that: The feedback module comprises: The text organization unit is used to combine the location result, time and situation of the item to be tracked into a text description.

Images