Event information merging method and device and event processing system

Abstract

At least one embodiment of the invention discloses an event information merging method, an event information merging device and an event processing system.The event information merging method comprises the steps that respective feature information of at least two to-be-processed events is obtained, and the feature information comprises attribute data related to event content; determining an association degree between the at least two to-be-processed events based on the feature information and a preset association standard, wherein the preset association standard comprises a system built-in standard and / or a user-defined association standard; and in response to the condition that the association degree meets a preset condition, combining the at least two to-be-processed events into one aggregation event.

Description

Technical Field

[0001] This disclosure relates to the field of data processing, and more specifically to a method for merging event information, an apparatus for merging event information, and an event processing system. Background Technology

[0002] Security systems are core technologies for ensuring public, industrial, and home safety. Event processing systems primarily use analog or digital cameras to capture video streams of the monitored area via infrared, visible light, and thermal imaging. However, with the surge in the number of videos captured by cameras and the increased frequency of events triggered by these cameras, users need to spend a significant amount of time and effort monitoring these events. Traditional security systems are gradually revealing a contradiction between the large volume of data and the user experience.

[0003] Security systems need to be improved to handle large volumes of events in order to enhance the user experience. Summary of the Invention

[0004] According to one aspect of this disclosure, an event information merging method is provided, comprising: acquiring feature information of at least two events to be processed, the feature information including attribute data related to the event content; determining the degree of correlation between the at least two events to be processed based on the feature information and a preset correlation standard, the preset correlation standard including a system-built-in standard and / or a user-defined correlation standard; and merging the at least two events to be processed into an aggregated event in response to the correlation degree meeting a preset condition.

[0005] In some embodiments, user-defined association criteria include at least one of the following: user-configured feature weight parameters, used to adjust the influence weight of different attribute data in the calculation of association degree; user-set event matching thresholds, used to define the criteria for determining the degree of association; and user-specified event attribute combination criteria, used to limit the type of feature information.

[0006] In some embodiments, the method further includes: providing a user interaction interface to receive custom parameters input by the user, wherein the custom parameters include feature selection instructions, weight allocation instructions, or threshold setting instructions, which are used to select the type of feature information, allocate feature weight parameters, or set an event matching threshold, respectively; storing the custom parameters as a user-personalized standard template and associating them with the system's built-in standards to form a hybrid association standard library.

[0007] In some embodiments, the feature information includes at least one of the following: the time attribute, spatial attribute, content attribute, source attribute, and / or associated object attribute of the event; wherein, the content attribute includes the text description of the event, semantic tags, and / or multimedia data features, and the source attribute includes the event's acquisition device identifier and / or data format.

[0008] In some embodiments, determining the degree of association between at least two events to be processed based on feature information and preset association criteria includes: comparing feature information using preset association criteria to obtain quantitative parameters or classification results that characterize the relevance of events as the basis for determining the degree of association; wherein, the quantitative parameters include similarity scores or association probability values.

[0009] In some embodiments, the preset conditions include at least one of the following: the correlation degree is higher than a preset threshold, the preset threshold including a system default threshold or a user-defined event matching threshold; the feature information of at least two events to be processed meets a preset matching standard, the matching standard including complete matching or partial matching; the user manually triggers a merging instruction based on the feature information.

[0010] In some embodiments, the method further includes: generating comprehensive information about the aggregated event, the comprehensive information including attributes of the aggregated event, a list of associated events, and / or summary data; and outputting the comprehensive information, the output of which includes displaying, storing, or transmitting the comprehensive information.

[0011] In some embodiments, if at least two events to be processed contain multimedia data, the aggregated information includes the result of integrating the multimedia data, including filtering, splicing, compression, or format conversion.

[0012] According to another aspect of this disclosure, an event information merging apparatus is provided, comprising: an information acquisition module configured to acquire feature information of at least two events to be processed, the feature information including attribute data related to the event content; an association analysis module configured to determine the degree of association between at least two events to be processed based on the feature information and a preset association standard, the preset association standard including system-built-in standards and / or user-defined association standards; and an event merging module configured to merge at least two events to be processed into an aggregated event in response to the degree of association meeting a preset condition.

[0013] In some embodiments, the apparatus further includes: a standard management module configured to store and retrieve preset associated standards, the preset associated standards including system-built-in standards and / or user-defined associated standards; wherein, the standard management module includes: a standard configuration unit configured to provide a user interaction interface, receive and parse user-input user-defined associated standards; and a standard storage unit configured to store system-built-in standards and / or user-personalized standard templates, and support adding, modifying or deleting personalized standard templates.

[0014] According to another aspect of this disclosure, an event information merging apparatus is provided, characterized in that the apparatus includes: at least one memory configured to store program instructions; and one or more processors coupled to the at least one memory and configured to execute the program instructions to perform a method according to at least one embodiment of this disclosure.

[0015] According to another aspect of this disclosure, an event processing system is provided, comprising: a data acquisition device configured to acquire video and event data; an event information merging apparatus according to at least one embodiment of this disclosure; and an output device configured to output aggregated event information to a user.

[0016] In some embodiments, the system further includes a visual configuration interface for users to configure user-defined association criteria through graphical operations. The user-defined association criteria include at least one of the following: user-configured feature weight parameters, used to adjust the influence weight of different attribute data in the calculation of the degree of association; user-set event matching thresholds, used to define the criteria for determining the degree of association; and user-specified event attribute combination criteria, used to limit the type of feature information.

[0017] According to at least one aspect of this disclosure, by determining the degree of correlation between at least two events to be processed based on feature information and preset correlation criteria, and in response to the degree of correlation meeting preset conditions, the at least two events to be processed are merged into one aggregated event. This can avoid the same event being captured and repeatedly triggered by multiple devices at the same time, or the same event being repeatedly triggered over time. It can also merge redundant and fragmented events, store and output only the merged event, avoid swallowing invalid information, improve information processing efficiency, and save storage space. Attached Figure Description

[0018] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to offer a further understanding of the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 A scene diagram of an event processing system according to at least one embodiment of the present disclosure is shown.

[0020] Figure 2 A flowchart of an event information merging method according to at least one embodiment of the present disclosure is shown.

[0021] Figure 3 A block diagram of an event information merging apparatus according to at least one embodiment of the present disclosure is shown.

[0022] Figure 4 A block diagram of another event information merging apparatus according to at least one embodiment of the present disclosure is shown. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this disclosure more apparent, exemplary embodiments according to this disclosure will now be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals denote the same elements throughout. It should be understood that the embodiments described in this disclosure are merely illustrative and should not be construed as limiting the scope of this disclosure.

[0024] The terms “exemplary,” “for example,” and / or “example” are used herein to mean “serving as an example, instance, or illustration.” Any aspect described herein as “exemplary,” “for example,” and / or “example” is not necessarily to be construed as superior to or better than other aspects. Similarly, the term “aspects of this disclosure” does not require that all aspects of this disclosure include the features, advantages, or modes of operation discussed.

[0025] Current security systems suffer from a contradiction between the sheer volume of events and the need for efficient user awareness. Traditional security cameras trigger over 100 events daily (such as motion detection and area intrusion). Current event processing mechanisms suffer from severe redundancy and fragmentation: the same event may be captured simultaneously by multiple security camera devices and repeatedly triggered, and the same event may be repeatedly triggered over time, resulting in a flood of useful information and low processing efficiency. Existing technologies often employ simple time filtering, failing to identify the essential relationships between events, leading to wasted storage resources and potential security vulnerabilities.

[0026] At least one embodiment of this disclosure aims to provide an event information merging method, an event information merging device, and an event processing system. By intelligently associating the characteristic information of events, it enables event merging across devices, across time, and / or across types, thereby improving event processing efficiency and accuracy, and further saving storage space.

[0027] Figure 1 A scene diagram of an event processing system 100 according to at least one embodiment of the present disclosure is shown.

[0028] The event processing system 100 may include a data acquisition device 101, an event information merging device 102, and an output device 103. The event processing system 100 may be used for security monitoring, etc.

[0029] The acquisition device 101 can acquire video and event data, such as real-time video data or video data input. It is adaptable to different security monitoring scenarios such as homes, supermarkets, and public safety, and generates related event data, such as event feature information generated from video data, including attribute data related to the event content. This attribute data includes, but is not limited to, the event's time attribute, spatial attribute, content attribute, source attribute, and / or associated object attribute.

[0030] The acquisition device 101 may include, for example, a camera. Cameras may include visible light cameras, infrared cameras, thermal imaging cameras, 3D cameras, ultraviolet cameras, and so on. There are many types of cameras, and selection can be made based on security scenario requirements (e.g., lighting conditions, accuracy requirements, concealment), technical characteristics (e.g., resolution, sensor type), and system integration capabilities (e.g., artificial intelligence (AI) algorithms, network transmission). The acquisition device 101 can also capture events from the acquired video data, such as identifying events occurring in the video data through an artificial intelligence recognition model, such as an elderly person falling, a child learning, etc., and collecting event data associated with these events.

[0031] The event information merging device 102 can perform data processing, such as acquiring feature information of at least two events to be processed, including attribute data related to the event content; determining the degree of correlation between at least two events to be processed based on the feature information and preset correlation standards, including system-built-in standards and / or user-defined correlation standards; merging at least two events to be processed into an aggregated event in response to the degree of correlation meeting preset conditions, and optionally generating comprehensive information of the aggregated event; outputting comprehensive information, etc.

[0032] The event information merging device 102 includes, for example, a processor and a memory. The processor may be a central processing unit (CPU), a graphics processing unit (GPU), a dedicated chip, a field-programmable gate array (FPGA), etc.

[0033] The output device 103 can receive the aggregated event information output by the event information merging device 102, and output the aggregated event information, and / or the aggregated video data contained in the aggregated event, etc. to the user.

[0034] Output device 103 may include a display device (e.g., a liquid crystal display (LCD) / organic light-emitting diode (OLED) display), a virtual reality (VR) / augmented virtual reality (AR) device, etc., to display aggregated information about the aggregated event. Output device 103 may also include a speaker for broadcasting to the user the sound of the aggregated information being read aloud or an alarm, etc.

[0035] System 100 may also include a visual configuration interface for users to configure user-defined association criteria through graphical operations. The user-defined association criteria may include at least one of the following: user-configured feature weight parameters, used to adjust the influence weight of different attribute data in the calculation of association degree; user-set event matching thresholds, used to define the criteria for determining the degree of association; and user-specified event attribute combination criteria, used to limit the type of feature information.

[0036] System 100 may also include other input devices, such as a touchscreen, keyboard / mouse, microphone, etc.

[0037] In practical applications, such as home security, users deploy multiple data acquisition devices, i.e., cameras. When a person enters the front yard, camera A triggers the identification of a "person intrusion" event, and camera B triggers the identification of a "person movement" event. The system determines the degree of correlation between the two events based on their respective event characteristics (e.g., the time attribute, spatial attribute, content attribute, source attribute, and / or associated object attribute). If the degree of correlation meets preset conditions, such as the location difference between the two events being less than 50 meters and the time interval being less than 30 seconds, the target similarity between the two events being higher than a predetermined threshold, or the semantic relevance between the two events being higher than a predetermined threshold, the system can merge the "person intrusion" event and the "person movement" event into a single aggregated event (e.g., person intrusion). The system can further generate and output comprehensive information about this aggregated event, including, for example, the attributes of the aggregated event (e.g., security event), a list of associated events (e.g., "person intrusion" event, "person movement" event), and / or summary data "December 10, 2025, 21:30, a person intruded into the front yard," and / or video clips related to the aggregated event.

[0038] Figure 2 A flowchart of an event information merging method 200 according to at least one embodiment of the present disclosure is shown.

[0039] like Figure 2 As shown, the event information merging method 200 includes steps 201, 202, and 203.

[0040] In step 201, feature information of at least two events to be processed is obtained, including attribute data related to the event content. In step 202, the degree of correlation between the at least two events to be processed is determined based on the feature information and preset correlation criteria, including system-built-in criteria and / or user-defined correlation criteria. In step 203, in response to the correlation degree meeting preset conditions, the at least two events to be processed are merged into one aggregated event.

[0041] Thus, according to at least one embodiment of this disclosure, by determining the degree of correlation between at least two events to be processed based on feature information and preset correlation criteria, and in response to the degree of correlation meeting preset conditions, the at least two events to be processed are merged into one aggregated event. This can prevent the same event from being captured and repeatedly triggered by multiple security camera devices at the same time, or the same event from being repeatedly triggered over time. It can also merge redundant and fragmented events, store and output only the merged event, avoid swallowing invalid information, improve information processing efficiency, and save storage space.

[0042] The above steps and related embodiments will now be described in detail.

[0043] In step 201, the feature information of at least two events to be processed is obtained. The feature information includes attribute data related to the event content.

[0044] In some embodiments, the feature information may include at least one of the following: time attribute, spatial attribute, content attribute, source attribute and / or associated object attribute of the event; wherein, the content attribute includes the text description of the event, semantic tags, and / or multimedia data features, and the source attribute includes the event's acquisition device identifier and / or data format.

[0045] The time attribute of an event can include the timestamp of the event and / or the duration. For example, in home security, the timestamp is 2025-12-10T21:30:00Z (ISO 8601 format), and the duration is 30 seconds (e.g., an "intrusion" event from 21:30:00 to 21:30:30), etc.; in supermarket security, the timestamp is 2025-12-10T14:45:00Z (weekday midday peak hours), and the duration is 5 minutes (e.g., from 14:45:00 to 14:50:00), etc.

[0046] Spatial attributes include the specific geographical location of the event or the location of the device, typically obtained via GPS or user-defined areas. For example, in home security, GPS coordinates might be 120.1234°E, 30.5678°N (camera A is installed in the front yard), and user-defined areas might include the front yard, living room, and bedroom (manually defined via a mobile application). In supermarket security, GPS coordinates might be 121.4737°E, 31.2304°N (camera B covers the beverage shelf area), and user-defined areas might include the entrance, cashier, and fresh produce section (preset via a management platform). Spatial attributes can also refer to the physical space where the event occurred, such as the living room area or "within 2 meters of the desk" (user-defined monitoring priority area) in home security; or the beverage shelf area or "within a 5-meter radius of the cashier" (user-defined sensitive area) in supermarket security.

[0047] The text descriptions or semantic tags for events in the content attributes can include automatically generated text information based on various data, or semantic tags obtained by semantically summarizing the event, making it easier for users to understand. For example, in home security, based on time information, content attributes, spatial attributes, target feature information, etc., the text description of the event "2025-12-10 21:30, a person in black stayed in the front yard for 5 minutes" (combining visual and action recognition results), or the event can be semantically summarized to generate the semantic tag "suspicious person loitering" (based on a comprehensive judgment of time, location, and behavior). In supermarket security, based on time information, content attributes, spatial attributes, target feature information, etc., the text description of the event "2025-12-10 14:45, beverage shelf goods were moved" (combining visual and action recognition results), and the event can be semantically summarized to generate the semantic tag "goods display adjustment," etc. Here, the automatically generated text descriptions or semantic tags of events can be used to train artificial intelligence models to obtain data from various real-world information by utilizing various information samples and text descriptions or semantic tag annotations.

[0048] Multimedia data features in content attributes can include visual information about events, such as video or image data captured by devices like cameras, containing visual content within a scene. For example, a home security camera might capture images of "a person in black loitering in the living room" or "a pet dog barking at the door." In supermarket security systems, surveillance footage might show images of "customers browsing merchandise on shelves" or "employees carrying goods to the checkout counter."

[0049] Multimedia data features in content attributes may also include event audio information, such as sound data collected by devices like microphones, containing ambient sounds or speech content. For example, in home security, ambient sounds such as microphone C detect "footsteps lasting 30 seconds" (classified by the model), and speech content such as "someone shouting 'open the door'" extracted by the speech recognition model (transcribed by the model); in supermarket security, ambient sounds such as microphone D recognize "frequent barcode scanner sounds in the cashier area," and speech content such as "customers asking 'when will the promotion end?'" extracted by the speech recognition model.

[0050] The source attribute is key information that describes the original source and characteristics of data or events. In scenarios involving the collection of multimedia data or events such as videos and audios, the source attribute usually includes the identification of the collection device and / or the data format of the collected data. Here, in addition to including the unique number of the collection device, the collection device identification can also include information about the installation location of the camera, such as floor, area, specific coordinates, etc. The data format specifies how the collected data is encoded, stored, and transmitted, such as AV1, MP4, TXT, and so on. Different data formats can be applicable to different application scenarios and devices.

[0051] The associated object attributes can include fine-grained feature information about the targets (such as people, vehicles, animals) involved in the event, which can be extracted through artificial intelligence (AI) models. For example, in home security, face information such as "the facial feature vector of Zhang San" recognized by camera A (extracted through a convolutional neural network (CNN) model), gait information such as "the target gait matches the 'child' feature in the known gait library" detected by camera B, clothing information including the target wearing a "red hoodie" (extracted through an image segmentation model to obtain color and style), etc.; in supermarket security, license plate information such as the vehicle feature being "Shanghai A12345" (extracted through an optical character recognition (OCR) model), item information, such as the target carrying a "white plastic bag containing beverage bottles" (recognized through an object detection model).

[0052] The above information describes the characteristics of the current event from multiple dimensions and can be collected through various sensing devices (such as cameras, microphones, GPS, clocks, etc.).

[0053] Of course, the above information is only an example, and more feature information of events can be collected according to specific security scenarios.

[0054] Note that feature information can also include multi-dimensional event features obtained through further processing of the original information or extracted by artificial intelligence models. Artificial intelligence models include, for example, person recognition models, age recognition models, action recognition models, gender recognition models, clothing recognition models, lighting recognition models, spatial recognition models, and so on. Multi-dimensional event features can include, but are not limited to, at least one of the following: object features, action features, location features, environmental features, time features, sound features, and comprehensive information features. For example, object features can include at least one of the following: person (e.g., elderly person, child), facial contour, age (e.g., 6 years old, adult, minor), gender (e.g., man, woman), clothing (e.g., loungewear, delivery uniform, takeout uniform), etc.; action features can include at least one of the following: movement trajectory (e.g., crossing from left to right, approaching, moving away), behavior pattern (e.g., reading, falling, walking), etc.; environmental features can include at least one of the following: geographical location (e.g., home, school, coffee shop), lighting conditions (e.g., daytime, nighttime), spatial layout (e.g., bedroom, living room), etc.; time features can include at least one of the following: year, month, day, hour, minute, weekday, season, etc.; sound features can include children's voices, text recognized from speech recognition, etc.; comprehensive information features can include text from comprehensive event information, etc. In addition to the features mentioned above, more multi-dimensional event features can be added according to specific security scenarios.

[0055] In scenarios such as home security, feature information may include the following:

[0056] . In supermarket security scenarios, for example, feature information may include the following:

[0057] .

[0058] Obtaining attribute data related to the event content (such as multi-dimensional event features) can significantly improve the accuracy of event merging. For example: spatial attributes + temporal attributes + associated object attributes: if the same target triggers "intrusion" events consecutively in adjacent cameras, the system can merge them into one event using GPS location (spatial), timestamp (time), and facial feature vector (target features). Or, associated object attributes + content attributes: if a camera detects "delivery person" and the content attribute is "delivery person injured" (event description), they can be merged into one event.

[0059] By fusing and analyzing feature information, more accurate event association and merging can be achieved, ultimately reducing redundant events and improving event processing efficiency and accuracy.

[0060] In step 202, the degree of correlation between at least two events to be processed is determined based on feature information and preset correlation criteria. Preset correlation criteria include system-built-in criteria and / or user-defined correlation criteria.

[0061] For example, in some embodiments, the built-in system criteria may include at least one of the following: spatial distance, time interval, target similarity, speech relevance, and semantic relevance, calculated based on the feature information of multiple events, and their respective weights, to determine the degree of association between the multiple events.

[0062] For example, for the time attribute of an event, the time interval between the time attributes of multiple events can be obtained. For example, if the time of event A is 10:00 AM and the time of event B is 10:05 AM, the time interval between the events is 5 minutes. It can be seen that the two times are very close, and the two events are more likely to be the same event that occurred at similar times.

[0063] For example, for spatial attributes, the spatial distance of multiple events can be obtained. For instance, the spatial distance between the location of event A and the location of event B is 5 meters. It can be seen that the two locations are very close, and the two events are more likely to be the same event in basically the same location.

[0064] For example, regarding associated object attributes, the similarity between associated object attributes of multiple events can be obtained. For instance, if the associated object attribute of event A is the facial contour of child A, and the associated object attribute of event B is also the facial contour of child A (which has a high target similarity), the two events are more likely to be the same event involving the same object. For example, when calculating the target similarity of multiple events, a convolutional neural network (such as MobileNet or FaceNet) can be used to extract the facial feature vector F1 of event A and the facial feature vector F2 of event B, and then calculate the cosine similarity. As a target similarity metric, the numerator is the dot product (inner product) of the vectors, reflecting directional consistency. The denominator is the product of the vector magnitudes, used for normalization to eliminate the influence of length.

[0065] For example, regarding content attributes, we can obtain the semantic or feature relevance between the content attributes of multiple events. For instance, if the content attribute of event A (the captured video content) is "people are walking," and the content attribute of event B is also "people are walking," a high semantic or feature relevance suggests that the two events are more likely to be recurring events related to the same walk. For example, when calculating semantic or feature relevance, we can use Natural Language Processing (NLP) models to convert the text descriptions or semantic tags of events into word vectors as semantic relevance, or calculate the cosine similarity of feature vectors composed of multimedia data features as feature relevance.

[0066] For example, regarding the source attribute of an event, the similarity between the source attributes of multiple events can be obtained. For instance, the source attribute of event A is, for example, camera A and data format AV1, and the source attribute of event B is, for example, camera A and data format AV1. The two are highly correlated. Combined with other information, the two events may be the same event in the same scene from the same camera.

[0067] Of course, the above calculation method can also include many other methods, which can be constructed based on the knowledge of those skilled in the art, and will not be described one by one here.

[0068] The system's built-in standard could be a weighted sum of spatial distance, time interval, relevance, and similarity information according to their respective weights to obtain the degree of association. Of course, this is just an example; other standards can also be used.

[0069] In some embodiments, user-defined association criteria include at least one of the following: user-configured feature weight parameters, used to adjust the influence weight of different attribute data in the calculation of association degree; user-set event matching thresholds, used to define the criteria for determining the degree of association; and user-specified event attribute combination criteria, used to limit the type of feature information.

[0070] For example, users can set weights for different attribute data to perform weighted summation when calculating correlation. For instance, a user might assign a higher weight to spatial attributes than others, perhaps because they don't want to receive event information from unrelated spatial locations. Users can specify that the correlation level must exceed a user-defined event matching threshold to be considered a valid event for merging at least two pending events into a single aggregate event. For example, a user might set a higher event matching threshold to avoid receiving frequent event notifications. Users can also limit which types of feature information (e.g., calculating only the correlation between time and spatial attributes) can be used to calculate correlation.

[0071] In some embodiments, the method further includes: providing a user interaction interface to receive user-inputted custom parameters, including feature selection instructions, weight allocation instructions, or threshold setting instructions (used respectively to select the type of feature information, allocate feature weight parameters, or set an event matching threshold); storing the custom parameters as a user-personalized standard template and associating them with the system's built-in standards to form a hybrid associative standard library. In this way, users can more flexibly set the relevant parameters for merging standards and can record various parameters set by the user each time for future reuse or as a recommended or default scheme.

[0072] In some embodiments, determining the degree of association between at least two events to be processed based on feature information and preset association criteria includes: comparing feature information using preset association criteria to obtain quantitative parameters or classification results that characterize the relevance of events as the basis for determining the degree of association; wherein, the quantitative parameters include similarity scores or association probability values.

[0073] For example, consider a security monitoring system for a large commercial complex. Multiple cameras are deployed across different floors and areas to monitor the activities of people and vehicles. The system needs to determine whether there is a spatial correlation between events captured by different cameras.

[0074] Feature information includes, for example, location information of the event captured by each camera, such as the specific floor and area coordinates. This location information constitutes the spatial attributes of the event.

[0075] During the comparison process, events A and B, captured by two different cameras, are obtained, and their location coordinates are extracted. If the GPS coordinates of event A are (lat1, lon1) and the GPS coordinates of event B are (lat2, lon2), then the spatial distance can be calculated using the following Euclidean distance formula: The spatial distance is calculated by measuring the Euclidean distance between the two coordinates, which can be used as a quantitative parameter (similarity score) to characterize the relevance of events. Alternatively, if the calculated spatial distance is 30 meters, less than a preset threshold of 50 meters, then events A and B are considered spatially close, and this spatial proximity is used as the classification result to characterize the relevance of events. If the spatial distance is 80 meters, greater than the threshold, then the two events are considered spatially distant, and this is used as the classification result to characterize the relevance of events (i.e., classified as close or distant). The spatial distance can also be further normalized to obtain a similarity score or association probability value between 0 and 1 (by rewriting it as a percentage). The closer the spatial distance, the closer the similarity score or association probability value is to 1, which is used as a quantitative parameter to characterize the relevance of events.

[0076] For example, a security monitoring system in a commercial complex needs to analyze events captured by the same camera at different times, or the temporal correlation of events captured by different cameras, in order to discover possible abnormal behavior patterns.

[0077] Feature information includes, for example, the timestamp of each event, which records the specific time when the event occurred. This time information constitutes the time attribute of the event.

[0078] During the comparison process, two events C and event D are obtained, and their timestamps are extracted respectively. Calculate the difference between the two timestamps to obtain the time interval. Specifically, if the timestamp of event C is Time1 and the timestamp of event D is Time2, the time interval can be calculated by the following formula: D time = |Time1 - Time2|. If the time interval value is 5 minutes, which is less than the preset threshold of 10 minutes, it is determined that events C and event D are close in time, and the time proximity is used as the classification result of the association degree; if the time interval value is 15 minutes, which is greater than the threshold, it is determined that the two events are not close in time. The time interval can also be converted to obtain a similarity score or an association probability value. For example, the shorter the time interval, the closer the similarity score or the association probability value is to 1, which is used as a quantitative parameter representing the event correlation.

[0079] Other examples are not described one by one here.

[0080] In some embodiments, the preset conditions include at least one of the following: the association degree is higher than a preset threshold, and the preset threshold includes a system default threshold or an event matching threshold set by the user; the feature information of at least two to-be-processed events meets a preset matching criterion, and the matching criterion includes a complete match (for example, all feature information meets the requirements) or a partial match (a part (for example, a certain percentage) of the feature information meets the requirements); the user manually triggers a merge instruction based on the feature information.

[0081] For example, the spatial distance between event A and event B is X1, the time interval is X2, the target similarity is X3, and the semantic correlation degree is X4. The association degree AD can be X1 * W1 + X2 * W2 + X3 * W3 + X4 * W4, and the weights W1, W2, W3, and W4 respectively correspond to their respective weights. The preset threshold can be AD threshold If AD > AD threshold then it is considered that the association degree meets the preset conditions, and at least two to-be-processed events are merged into one aggregated event. Of course, the above examples are only examples and not limitations, and other examples can also be constructed.

[0082] Another example is that the preset conditions can include a part of the feature information matching requirements, such as the spatial distance < T1 and the time interval < T2, but other feature information may not meet the requirements (i.e., the case of partial match), or include the spatial distance between event A and event B < T1 and the time interval < T2, and the target similarity > T3, and the semantic correlation degree > T4 (i.e., the case of complete match). Then it is considered that the association degree meets the preset conditions, and at least two to-be-processed events are merged into one aggregated event. Of course, the above examples are only examples and not limitations, and other examples can also be constructed.

[0083] For example, a user might want to merge events according to their own requirements. For instance, a user might want to merge events that meet their specified content attributes of "name" and "time." For example, a user might want to aggregate events related to "Mark today," in which case they could merge multiple events whose content attributes include "Mark" and whose time attribute includes "today." Or, a user might want to merge all events related to cats using the litter box, in which case they could merge multiple events whose content attributes include "cat" and whose location attribute includes "litter." Of course, the above examples are merely illustrations and not limitations; other examples can be constructed.

[0084] There may be other features, degree of association, preset association standards, and preset conditions, which will not be described in detail here.

[0085] In some embodiments, method 200 may further include: generating comprehensive information of the aggregated event, the comprehensive information including attributes of the aggregated event, a list of associated events and / or summary data; and outputting the comprehensive information, the output of which includes displaying, storing or transmitting the comprehensive information.

[0086] For example, in the scenario of a smart security system in a bank branch, the aggregated events are classified as low-risk (normal business behavior).

[0087] The list of associated events is as follows.

[0088] .

[0089] The summary data is: "Person P1 entered the bank branch through camera A at 10:05:30, then lingered in front of the counter for 3 minutes and 30 seconds at 10:08:15, completed the withdrawal operation at ATM C at 10:12:40, and left the branch through camera D at 10:15:20."

[0090] In this way, what used to require processing four independent events can now be viewed by users in a single aggregated event summary, saving 80% of playback time. Through automatic association and summary generation, users can quickly identify normal or abnormal behavior patterns.

[0091] In some embodiments, if at least two events to be processed contain multimedia data, the aggregated information includes the result of integrating the multimedia data, including filtering, splicing, compression, or format conversion.

[0092] For example, a smart security system in a bank branch detects two related events (Event 1 and Event 2) using multiple cameras and audio devices, each containing different multimedia data (video, audio, images, etc.). The system merges them into a single aggregated event based on their correlation and then filters, splices, compresses, and converts the multimedia data to generate comprehensive information.

[0093] For example, filtering could include extracting key segments from the video of Event 1: the scene of the customer inserting their bank card and entering their PIN (5 seconds), and extracting the scene of the customer leaving the ATM from the video of Event 2 (3 seconds). The audio of Event 1 (ATM operation sounds) and the image of Event 2 (front view) would be retained.

[0094] The splicing process can include stitching together selected video clips in chronological order to form a single integrated video, with the timeline as follows: 10:05:00-10:05:05 (withdrawal action) + 10:05:15-10:05:18 (departure action), totaling 8 seconds in length. Audio and video synchronization: The ATM operation audio is embedded into the integrated video (10:05:05-10:05:10).

[0095] Compression can include using H.265 encoding on the integrated video, reducing the file size from the original 500MB to 150MB. For images, JPEG compression can reduce the resolution from 1920×1080 to 1080×720, decreasing the file size from 2MB to 500KB.

[0096] Format conversion can include converting integrated video to WebM format to meet the fast loading requirements of web pages, or, for example, converting audio to AAC format for easy cross-platform playback.

[0097] In some embodiments, the result of integrating multimedia data included in the aggregated information can result in a data volume that is less than or equal to the total data volume of video data corresponding to at least two events. By performing the aforementioned processing, for example, making the aggregated video data volume less than the total data volume of video data corresponding to at least two events, it is possible to reduce the time and effort users spend watching redundant events, as well as reduce the storage space and time costs associated with hardware storage and transmission of redundant events.

[0098] This aggregated information can then be output to a display for the user to view, select, or modify, or output to a speaker for the speaker to read aloud. This allows only the aggregated events to be displayed, reducing the number of events the user needs to view, decreasing video playback time, and improving the user experience.

[0099] In summary, by intelligently associating event feature information, events can be merged across devices, time periods, and / or types, thereby improving event processing efficiency and accuracy, and further saving storage space.

[0100] Figure 3 A block diagram of an event information merging apparatus 300 according to at least one embodiment of the present disclosure is shown.

[0101] like Figure 3 As shown, the event information merging device 300 includes an information acquisition module 301, a correlation analysis module 302, and an event merging module 303.

[0102] The information acquisition module 301 is configured to acquire feature information of at least two events to be processed, including attribute data related to the event content.

[0103] The association analysis module 302 is configured to determine the association relationship between multiple events based on the feature information of multiple events.

[0104] The event merging module 303 is configured to determine the degree of association between at least two events to be processed based on feature information and preset association criteria. The preset association criteria include system built-in criteria and / or user-defined association criteria.

[0105] In some embodiments, the event merging module 303 is further configured to merge at least two pending events into one aggregated event in response to a preset condition being met in terms of correlation.

[0106] In some embodiments, the device 300 further includes: a standard management module configured to store and retrieve preset associated standards, the preset associated standards including system-built-in standards and / or user-defined associated standards; wherein, the standard management module includes: a standard configuration unit configured to provide a user interaction interface, receive and parse user-input user-defined associated standards; and a standard storage unit configured to store system-built-in standards and / or user-personalized standard templates, and support adding, modifying or deleting personalized standard templates.

[0107] Other details of the device and the technical effects it can achieve can be found in the details of the methods described above, and will not be repeated here.

[0108] Figure 4 A block diagram of another event information merging apparatus 400 according to at least one embodiment of the present disclosure is shown.

[0109] The event information merging apparatus 400 includes at least: at least one memory 420 configured to store program instructions; and one or more processors 410 coupled to the at least one memory and configured to execute program instructions to perform the event information merging method described above in at least one embodiment of this disclosure.

[0110] For example, only one processor 410 is shown, but there can also be multiple processors. Furthermore, processing can be performed by a single processor, or by more than one processor simultaneously, sequentially, or using other methods. Additionally, processor 410 can be mounted on more than one chip.

[0111] The functions of the event information merging device 400 can be implemented, for example, by reading the instructions (programs) stored in the memory 430 into hardware such as the processor 410 and the memory 420, so that the processor 410 can perform operations, control the communication performed by the communication device 430, and control the reading and / or writing of data in the memory 420.

[0112] Processor 410 enables the operating system to operate, thereby controlling the device as a whole. Processor 410 may be composed of a central processing unit (CPU) that includes interfaces with peripheral devices, control devices, arithmetic devices, registers, etc. For example, the aforementioned processing units can be implemented by processor 410.

[0113] Furthermore, the processor 410 reads programs (program code), data, etc., from the memory 420 and performs various processes accordingly. The program can be one that causes the computer to perform at least a portion of the actions described in the above embodiments. For example, the method executed by the event information merging device can be implemented using a control program stored in the memory 420 and operated by the processor 410.

[0114] The memory 420 may be a computer-readable recording medium, such as at least one of a read-only memory (ROM), an erasable programmable read-only memory (EPROM), an electrically programmable read-only memory (EEPROM), a random access memory (RAM), or other suitable storage media. The memory 420 may include registers, caches, main memory (main storage device), etc. The memory 420 may store executable programs (program code), software devices, etc., for implementing the methods according to an embodiment of this disclosure.

[0115] In addition, the memory 420 may also include a computer-readable recording medium comprising, for example, at least one of a flexible disk, a floppy disk, a magneto-optical disk (e.g., a CD-ROM (Compact Disc ROM), a Digital Universal Optical Disc, a Blu-ray disc), a removable disk, a hard disk, a smart card, a flash memory device (e.g., a card, a stick, a key driver), a magnetic stripe, a database, a server, or other suitable storage media.

[0116] Communication device 430 is hardware (transmitting and receiving device) used for communication between computers via wired and / or wireless networks, and is also referred to as a network device, network controller, network interface card (NIC), communication device, etc. To implement, for example, frequency division duplex (FDD) and / or time division duplex (TDD), communication device 430 may include high-frequency switches, duplexers, filters, frequency synthesizers, etc. For example, the aforementioned transmitting and receiving operations can be implemented by communication device 430.

[0117] The communication device 430 can communicate bidirectionally via one or more antennas, wired or wireless links, as described herein. For example, the communication device 430 can represent a wireless transceiver and can communicate bidirectionally with another wireless transceiver. The communication device 430 may also include a modem to modulate packets and provide the modulated packets to the antenna for transmission, and to demodulate packets received from the antenna. In some cases, the communication device may include a single antenna 450. However, in some cases, the communication device may have more than one antenna 450, which are capable of simultaneously transmitting or receiving multiple wireless transmissions.

[0118] Furthermore, the processor 410, memory 420, and other devices are connected via a bus 440 for communication of information. The bus 440 can consist of a single bus or different buses between devices.

[0119] Furthermore, the devices and systems mentioned in this article may include hardware such as microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), programmable logic devices (PLDs), and field-programmable gate arrays (FPGAs).

[0120] This disclosure may include a non-transitory computer-readable storage medium. Instructions, such as computer instructions, are stored on the non-transitory computer-readable storage medium. When the computer instructions are executed by a processor, the various methods described above can be performed. Non-transitory computer-readable storage media include, but are not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, computer storage media (e.g., hard disks, floppy disks, solid-state drives, removable disks, CD-ROMs, DVD-ROMs, Blu-ray discs, etc.). For example, the non-transitory computer-readable storage medium can be connected to a computing device such as a computer, and then, when the computing device executes the computer instructions stored on the computer-readable storage medium, the various methods described above can be performed.

[0121] This disclosure may also include a computer program product capable of performing the methods, steps, and operations given herein. For example, such a computer program product may be a computer software package, computer code instructions, or a computer-readable tangible medium having computer instructions tangibly stored (and / or encoded) thereon, which can be executed by a processor to perform the operations described herein. The computer program product may include packaging materials.

[0122] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The term “such as / for example” as used herein refers to the phrase “such as / for example but not limited to,” and is used interchangeably with it.

[0123] The flowcharts and method descriptions in this disclosure are merely illustrative examples and are not intended to require or imply that the steps of the various embodiments must be performed in the given order. As those skilled in the art will recognize, the steps in the above embodiments can be performed in any order. Words such as "then," "next," etc., are not intended to limit the order of the steps; these words are only used to guide the reader through the description of these methods. Furthermore, any reference to a singular element, such as the use of the articles "a," "one," or "the," is not to be construed as limiting that element to the singular.

[0124] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit at least one embodiment of the present disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for merging event information, characterized in that, include: Obtain feature information for at least two events to be processed, the feature information including attribute data related to the event content; The degree of correlation between the at least two events to be processed is determined based on the feature information and the preset correlation criteria, wherein the preset correlation criteria include system built-in criteria and / or user-defined correlation criteria. If the correlation degree meets the preset conditions, the at least two pending events are merged into one aggregated event.

2. The method according to claim 1, characterized in that, The user-defined association criteria include at least one of the following: User-configured feature weight parameters are used to adjust the influence weight of different attribute data in the calculation of the degree of association; The user-defined event matching threshold is used to define the criteria for determining the degree of association. The user-specified event attribute combination criteria are used to define the type of the feature information.

3. The method according to claim 2, characterized in that, The method further includes: Provide a user interaction interface to receive custom parameters input by the user. The custom parameters include feature selection instructions, weight allocation instructions, or threshold setting instructions, which are used to select the type of feature information, allocate the feature weight parameters, or set the event matching threshold, respectively. The custom parameters are stored as user-personalized standard templates and associated with the system's built-in standards to form a hybrid association standard library.

4. The method according to claim 1, characterized in that, The feature information includes at least one of the following: The event's time attribute, location attribute, content attribute, source attribute, and / or associated object attribute; The content attributes include the text description, semantic tags, and / or multimedia data features of the event, and the source attributes include the acquisition device identifier and / or data format of the event.

5. The method according to claim 1, characterized in that, Determining the degree of correlation between the at least two events to be processed based on the feature information and the preset correlation criteria includes: The feature information is compared using the preset association criteria to obtain quantitative parameters or classification results that characterize the relevance of events, which serve as the basis for determining the degree of association. The quantification parameters include similarity scores or association probability values.

6. The method according to claim 1, characterized in that, The preset conditions include at least one of the following: The correlation degree is higher than a preset threshold, which includes a system default threshold or a user-defined event matching threshold; The feature information of at least two events to be processed satisfies a preset matching standard, which includes complete matching or partial matching. The user manually triggers the merge command based on the aforementioned feature information.

7. The method according to claim 1, characterized in that, The method further includes: Generate comprehensive information about the aggregated event, which includes the attributes of the aggregated event, a list of related events, and / or summary data. Output the comprehensive information, which includes displaying, storing, or transmitting the comprehensive information.

8. The method according to claim 7, characterized in that, If the at least two events to be processed contain multimedia data, the integrated information includes the result of integrating the multimedia data, the integration process including filtering, splicing, compression, or format conversion.

9. An event information merging device, characterized in that, include: The information acquisition module is configured to acquire feature information of at least two events to be processed, the feature information including attribute data related to the event content; The correlation analysis module is configured to determine the degree of correlation between the at least two events to be processed based on the feature information and preset correlation criteria, wherein the preset correlation criteria include system built-in criteria and / or user-defined correlation criteria. The event merging module is configured to merge the at least two pending events into one aggregated event in response to the correlation degree meeting a preset condition.

10. The apparatus according to claim 9, characterized in that, The device further includes: The standard management module is configured to store and retrieve the preset association standards, which include system-built-in standards and / or user-defined association standards; The standard management module includes: The standard configuration unit is configured to provide a user interaction interface, receiving and parsing user-defined association standards input by the user. The standard storage unit is configured as a built-in standard and / or a user-customized standard template of the storage system, and supports adding, modifying or deleting the personalized standard template.

11. An event information merging device, characterized in that, The device includes: At least one memory is configured to store program instructions; and One or more processors, said one or more processors coupled to said at least one memory, and configured to execute said program instructions to perform the method of any one of claims 1 to 8.

12. An event processing system, characterized in that, include: The acquisition device is configured to acquire video and event data; The event information merging apparatus according to any one of claims 9-11; The output device is configured to output comprehensive information about the aggregated events to the user.

13. The system according to claim 12, characterized in that, The system also includes a visual configuration interface, which allows users to configure user-defined association criteria through graphical operations. The user-defined association criteria mentioned above include at least one of the following: User-configured feature weight parameters are used to adjust the influence weight of different attribute data in the calculation of the degree of association; The user-defined event matching threshold is used to define the criteria for determining the degree of association. The user-specified event attribute combination criteria are used to define the type of the feature information.

Images