A Key Object Recognition Method and System Based on Border Object Relationships

Abstract

The present invention discloses a key object recognition method and system based on border object relationships, which relates to the technical field of data processing. The method includes: obtaining historical event information and obtaining an associated event object set based on the historical event information; obtaining monitoring data and performing frequent item set mining to obtain a final frequent item set; obtaining a monitoring object set and its corresponding time and location based on the monitoring data; obtaining a relationship set of all events based on spatio-temporal rules based on the monitoring object set and the time and location; generating an object relationship graph based on the associated event object set, the final frequent item set, and the relationship set; obtaining an event level weight, an object relationship weight, and an occurrence frequency weight based on the object relationship graph; obtaining a comprehensive score based on the event level weight, the object relationship weight, and the occurrence frequency weight; and determining key objects based on the comprehensive score. It can accurately identify key objects and improve the recognition efficiency of key objects.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a method and system for identifying key objects based on border object relationships. Background Art

[0002] Currently, in the field of border control and management, based on border event information and monitoring data from border monitoring equipment and facilities, by integrating the feature data of objects in the events and monitored objects, a relationship analysis model of border objects is established to identify key objects in the object relationships, which is of great help for discovering border event relationships and improving border control efficiency.

[0003] However, traditional association analysis algorithms are mainly applied in fields such as urban public security prevention and control, finance, technology, and healthcare, and are less applied in the field of border control and management. Moreover, generally, a single support degree algorithm is used, which cannot meet the diverse characteristics of border objects and cannot accurately identify key objects in events.

[0004] Therefore, how to accurately identify key objects and improve the identification efficiency of key objects is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a method and system for identifying key objects based on border object relationships, which can accurately identify key objects and improve the identification efficiency of key objects.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for identifying key objects based on border object relationships, comprising:

[0008] Obtaining historical event information and obtaining an associated event object set based on the historical event information;

[0009] Obtaining monitoring data and performing frequent item set mining to obtain a final frequent item set;

[0010] Obtaining a monitoring object set and its corresponding time and location based on the monitoring data;

[0011] Obtaining a relationship set of all events based on spatio-temporal rules based on the monitoring object set and the time and location;

[0012] Generating an object relationship graph based on the associated event object set, the final frequent item set, and the relationship set;

[0013] Obtaining an event level weight, an object relationship weight, and an appearance frequency weight based on the object relationship graph;

[0014] Obtain a comprehensive score based on the event level weight, the object relationship weight, and the occurrence frequency weight;

[0015] Determine key objects based on the comprehensive score.

[0016] Preferably, the method for obtaining the set of associated event objects is as follows:

[0017] Obtain an association table of border event object information including event identifiers and event object identifiers based on the historical event information;

[0018] Statistically obtain sets of objects corresponding to multiple different events as the set of event objects based on the association table of border event object information;

[0019] Merge the sets of event objects that are associated based on the association relationship of border events as the set of associated objects to obtain the set of associated event objects.

[0020] Preferably, the method for obtaining the final frequent item set is as follows:

[0021] Obtain a set of surveillance images for multiple frames based on the surveillance data;

[0022] Obtain a set of surveillance objects based on all the surveillance objects appearing in each frame of the surveillance image;

[0023] Obtain multiple sub-frequent item sets based on the set of surveillance objects;

[0024] Obtain the final frequent item set by taking the union of all the sub-frequent item sets.

[0025] Preferably, obtaining multiple sub-frequent item sets based on the set of surveillance objects specifically includes:

[0026] S201 Traverse each surveillance image in sequence based on the set of surveillance objects;

[0027] S202 Obtain the k-item set of the set of surveillance objects based on the number of surveillance objects in the surveillance image and the current frequent item number k;

[0028] S203 Obtain the sub-frequent item set corresponding to the frequent item number k based on the support degree of each item in the k-item set;

[0029] S204 Based on the frequent item number k plus 1, repeat the above S201 - S203 until no new frequent items are generated, and obtain multiple sub-frequent item sets corresponding to different frequent item numbers.

[0030] Preferably, the method for obtaining the k-item set is as follows:

[0031] Compare the number m of monitored objects in the monitored image with the current frequent item number k;

[0032] If m > k, then obtain multiple k-item sets based on the permutation and combination of k monitored objects in the current monitored image;

[0033] If m = k, then use all the monitored objects in the current monitored image as one k-item set;

[0034] If m < k, then do not generate the k-item set;

[0035] Traverse all the monitored images, and all the obtained k-item sets form the k-item set collection.

[0036] Preferably, the method for obtaining the set of relationships of all events based on spatio-temporal rules is as follows:

[0037] Traverse each monitored object in the monitored object set;

[0038] Based on the current monitored object, successively judge the spatio-temporal coincidence with all other monitored objects in the monitored object set based on the time and location;

[0039] When the spatio-temporal coincidence condition is met, there is an association relationship, and all the association sets of the current monitored object are obtained;

[0040] Otherwise, there is no association relationship;

[0041] After traversing all the monitored objects, all the obtained association sets are used as the set of relationships of all events based on spatio-temporal rules.

[0042] Preferably, the method for generating the object relationship graph is as follows:

[0043] Create an initial relationship graph;

[0044] Traverse and judge whether the event objects in all the associated object sets in the associated event object set exist in the initial relationship graph;

[0045] If not, add all the event objects in the current associated object set to the initial relationship graph, and connect the event objects pairwise to obtain the first relationship graph;

[0046] Traverse and judge whether the monitored objects in all the sub-frequent item sets in the final frequent item set exist in the first relationship graph;

[0047] If not, add all the monitored objects in the current sub-frequent item set to the first relationship graph, and connect the monitored objects pairwise to obtain the second relationship graph;

[0048] Traverse and determine whether all the monitored objects in the associated sets in the relationship set exist in the second relationship graph;

[0049] If not, add all the monitored objects in the current associated set to the second relationship graph, and connect the monitored objects pairwise to obtain the object relationship graph;

[0050] During the above judgment process, if they exist, no operation is performed.

[0051] Preferably, the method for obtaining the event level weight and the object relationship weight is as follows:

[0052] All events in the object relationship graph correspond to corresponding level weights;

[0053] Traverse all the events corresponding to each event object in the object relationship graph, and assign the highest level weight corresponding to all the events to the corresponding event object as the event level weight;

[0054] Based on the number of connections corresponding to the monitored object with the largest number of connections in the object relationship graph as the maximum number of connections;

[0055] Traverse all the monitored objects in the object relationship graph, and use the ratio of the number of connections of the monitored object to the maximum number of connections as the object relationship weight.

[0056] Preferably, it further includes: obtaining the key area organization based on the key objects, specifically:

[0057] S901 Create a new organization, and use the unmarked key objects in the object relationship graph as the current key objects;

[0058] S902 Select the objects that have a connection relationship with the current key object and the number of intermediate connection objects is less than the set value from the object relationship graph and add them to the new organization, and mark the current key object;

[0059] S903 Judge whether there are unmarked key objects based on the new organization;

[0060] S904 If there are, select the objects that have a connection relationship with the unmarked key object and the number of intermediate connection objects is less than the set value from the object relationship graph and add them to the new organization, and mark the unmarked key object; if not, execute S6;

[0061] S905 Repeat S903 - S904 until all the key objects in the new organization are marked;

[0062] Repeat the execution of S901 - S905 until all the key objects in the object relationship graph are classified into the corresponding organizations, obtaining multiple key area organizations.

[0063] A key object recognition system based on border object relationships, comprising: a first data processing module, a second data processing module, a third data processing module, a relationship graph generation module, a graph processing module, and a key object recognition module;

[0064] The first data processing module is used to obtain historical event information and obtain an associated event object set based on the historical event information;

[0065] The second data processing module is used to obtain monitoring data and perform frequent item set mining to obtain a final frequent item set;

[0066] The third data processing module is used to obtain a monitoring object set and its corresponding time and location based on the monitoring data; obtain a relationship set of all events based on spatio - temporal rules based on the monitoring object set and the time and location;

[0067] The relationship graph generation module is used to generate an object relationship graph based on the associated event object set, the final frequent item set, and the relationship set;

[0068] The graph processing module is used to obtain an event level weight, an object relationship weight, and an appearance frequency weight based on the object relationship graph;

[0069] The key object recognition module is used to obtain a comprehensive score based on the event level weight, the object relationship weight, and the appearance frequency weight; determine key objects based on the comprehensive score.

[0070] It can be seen from the above - mentioned technical solutions that, compared with the prior art, the present invention discloses a key object recognition method and system based on border object relationships, having the following beneficial effects:

[0071] 1. The present invention proposes a key object recognition method for a border object relationship graph based on event levels and monitoring object frequencies. The relationship graph makes the key object recognition interpretable, is conducive to visual display, and thus improves the key object recognition efficiency.

[0072] 2. The algorithms of the present invention for performing frequent item calculations based on monitoring data and analyzing object relationships based on spatio - temporal rules can scientifically reflect the relationship set between objects, and thus can accurately identify key objects.

[0073] 3. The present invention calculates the importance weight of an object by comprehensively considering factors such as the level of comprehensive events, object relationships, and occurrence frequencies. The calculation results can reflect the importance of an object in the relationship graph, and the object set is organized and divided based on key objects, enabling efficient and accurate acquisition of border object organizations. Description of the Drawings

[0074] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on the provided drawings without creative efforts.

[0075] Figure 1 It is a flowchart of a key object recognition method based on border object relationships provided by the present invention.

[0076] Figure 2 It is a flowchart of a method for obtaining multiple sub-frequent item sets based on a monitoring object set provided by the present invention.

[0077] Figure 3 It is a flowchart of an object relationship graph generation method provided by the present invention.

[0078] Figure 4 It is a schematic structural diagram of a key object recognition system based on border object relationships provided by the present invention. Detailed Embodiments

[0079] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0080] Embodiment 1

[0081] As Figure 1 shown, the embodiment of the present invention discloses a key object recognition method based on border object relationships, including:

[0082] Obtain historical event information and obtain an associated event object set based on the historical event information;

[0083] Obtain monitoring data and perform frequent item set mining to obtain the final frequent item set;

[0084] Obtain a monitoring object set and its corresponding time and location based on the monitoring data;

[0085] Obtain the set of relationships based on the spatio-temporal laws of all events based on the set of monitored objects and time and location;

[0086] Generate an object relationship graph based on the set of associated event objects, the final frequent item set, and the set of relationships;

[0087] Obtain the event level weight, object relationship weight, and occurrence frequency weight based on the object relationship graph;

[0088] Obtain a comprehensive score based on the event level weight, object relationship weight, and occurrence frequency weight;

[0089] Determine key objects based on the comprehensive score.

[0090] Embodiment 2

[0091] An embodiment of the present invention discloses a method for identifying key objects based on border object relationships, including:

[0092] Obtain historical event information and obtain a set of associated event objects based on the historical event information.

[0093] Preferably, the method for obtaining the set of associated event objects is:

[0094] Obtain a border event object information association table including event identifiers and event object identifiers based on the historical event information;

[0095] Statistically obtain the set of objects corresponding to multiple different events as the set of event objects based on the border event object information association table;

[0096] Based on the association relationship of border events, merge the associated set of event objects as the set of associated objects to obtain the set of associated event objects.

[0097] Preferably, in this embodiment, border objects such as people, vehicles, ships, livestock, etc. in each border event are obtained based on the obtained historical event information;

[0098] Based on the correspondence between border objects and border events, obtain event object identifiers and event identifiers, and form a border event object information association table.

[0099] Preferably, the border event object information association table , , represents the nth record of the object information association table, including the event identifier and the objects that appear in this event; count all event objects in each event to form a set of objects under the same event , where C m represents the set of event objects corresponding to the mth event.

[0100] Preferably, the border event information includes the event name, event level, and event time.

[0101] Preferably, based on the association relationship of border events, the event object sets corresponding to the associated events are merged, indicating that the event objects in multiple associated events are related.

[0102] Obtain monitoring data and perform frequent item set mining to obtain the final frequent item set.

[0103] Preferably, in the analysis of border object relationships, in addition to being able to use the above historical event information data for object analysis, monitoring data can also be used for object relationship analysis. By obtaining the objects (people, vehicles, livestock, ships, etc.) that appear simultaneously in the monitoring images or videos and performing frequent item set mining on them, the relationships between objects can be analyzed.

[0104] Preferably, the method for obtaining the final frequent item set is as follows:

[0105] Obtain a set of monitoring images for multiple frames based on the monitoring data;

[0106] Based on all the monitoring objects that appear in each frame of the monitoring image, obtain a monitoring object set;

[0107] Obtain multiple sub-frequent item sets based on the monitoring object set;

[0108] Take the union of all the sub-frequent item sets to obtain the final frequent item set.

[0109] Preferably, as Figure 2 shown, obtaining multiple sub-frequent item sets based on the monitoring object set specifically includes:

[0110] S201 Traverse each monitoring image in turn based on the monitoring object set;

[0111] S202 Based on the number of monitoring objects in the monitoring image and the current frequent item number k, obtain the k-item set of the monitoring object set;

[0112] S203 Based on the support degree of each item in the k-item set, obtain the sub-frequent item set corresponding to the frequent item number k;

[0113] S204 Use the frequent item number k + 1 as the new current frequent item number, repeat the above S201 - S203, and obtain the sub-frequent item sets corresponding to different frequent item numbers of the monitoring object set until no new frequent items are generated, and obtain multiple sub-frequent item sets corresponding to different frequent item numbers.

[0114] Preferably, the initial value of the frequent item number k in this embodiment is 2.

[0115] Preferably, based on the monitoring object set , traverse each monitoring image T in turn n; obtaining the k-item set R of the monitoring object set based on the number m of monitoring objects in the monitoring image and the current frequent item number k k , and the number of occurrences w of each item in the k-item set R k , where the number of occurrences w indicates that there are w monitoring images in the monitoring object set T that include R k .

[0116] Preferably, the method for obtaining the k-item set is as follows:

[0117] Comparing the number m of monitoring objects in the monitoring image with the current frequent item number k;

[0118] If m > k, then multiple k-item sets are obtained by permuting and combining k monitoring objects in the current monitoring image;

[0119] If m = k, then all monitoring objects in the current monitoring image are used as a k-item set;

[0120] If m < k, then no k-item set is generated;

[0121] Traverse all monitoring images, and all obtained k-item sets form a k-item set.

[0122] Preferably, when m > k, the following are obtained by permuting and combining k monitoring objects in the current monitoring image: k-item sets.

[0123] Preferably, obtaining the sub-frequent item set corresponding to the frequent item number k based on the support degree of each k-item set in the k-item set includes:

[0124] Based on the k-item set R k solving the support degree for each k-item set A in it :

[0125] ;

[0126] Among them, represents the number of times A appears in R k , represents the total number of frequent items in R k , is a preset value, generally not exceeding 10; usually, if a person appears in a vehicle a certain number of times, then there must be a relationship between the vehicle and him;

[0127] Comparing the obtained support degree with the minimum support degree , if the support degree is greater than the minimum support degree then it indicates that the k-item set corresponding to this support degree is a sub-frequent item set, otherwise it does not belong.

[0128] Preferably, the minimum support degree is a preset value. The larger k is, the smaller.

[0129] Based on the monitoring data, a set of monitored objects and their corresponding time and location are obtained.

[0130] Preferably, according to the monitoring data, a set of monitored objects can be obtained , representing the s-th monitored object, as well as the appearance time and location of each monitored object.

[0131] Based on the set of monitored objects and the time and location, a set of relationships of all events based on spatio-temporal rules is obtained.

[0132] Preferably, the method for obtaining the set of relationships of all events based on spatio-temporal rules is as follows:

[0133] Traverse each monitored object in the set of monitored objects;

[0134] Based on the current monitored object, successively perform spatio-temporal coincidence judgment with all other unjudged monitored objects in the set of monitored objects based on the time and location;

[0135] When the spatio-temporal coincidence condition is met, there is an association relationship, and all association sets of the current monitored object are obtained;

[0136] Otherwise, there is no association relationship;

[0137] After traversing all the monitored objects, all the association sets are obtained as the set of relationships of all events based on spatio-temporal rules.

[0138] Preferably, traverse the set of monitored objects, is the currently traversed monitored object, j = 1, is the unjudged monitored object, h = 2.

[0139] Based on and 's appearance time and location, perform spatio-temporal coincidence judgment. If the spatio-temporal coincidence condition is met, then the objects and are associated; otherwise, they are not associated;

[0140] Let h = h + 1, loop the above steps, and finally obtain 's all association sets;

[0141] Let j = j + 1, h = j + 1, loop the above steps, and finally obtain the association sets of all monitored objects as the set of relationships of all events based on spatio-temporal rules S.

[0142] Preferably, according to the monitoring data, object combinations that occur simultaneously and regularly within a certain period in the same area are counted. For example, after object a appears for a certain period of time, object b also appears in the same area. If this situation occurs a certain number of times, there is a certain relationship between object a and object b.

[0143] In the border monitoring data, in addition to the objects that appear simultaneously may be related, objects that appear at different times and locations may also have potential connections. Based on the above, the spatio-temporal coincidence condition in this embodiment is specifically:

[0144] If within a certain time interval after monitoring object a appears, monitoring object b appears at the same location, and this phenomenon occurs times, then monitoring objects a and b are related;

[0145] If monitoring object a appears at a certain fixed time point in a day at a place, and monitoring object b appears at another fixed time point in a day at the same place, and this phenomenon occurs times, then monitoring objects a and b are related;

[0146] If monitoring object a appears at multiple locations within a period of time g, and monitoring object b also appears at these places during this period of time g, then these monitoring objects a and b are related;

[0147] If the appearance time and location of two monitoring objects meet any one of the above spatio-temporal coincidence conditions, it means there is an associated relationship.

[0148] Generate an object relationship graph based on the associated event object set, the final frequent item set, and the relationship set.

[0149] Preferably, as Figure 3 shown, the method for generating an object relationship graph is:

[0150] Create an initial relationship graph;

[0151] Traverse and determine whether the event objects in all associated object sets in the associated event object set exist in the initial relationship graph;

[0152] If not, add all event objects in the current associated object set to the initial relationship graph, and connect the event objects pairwise to obtain the first relationship graph;

[0153] Traverse and determine whether the monitored objects in all sub-frequent item sets in the final frequent item set exist in the first relationship graph;

[0154] If not, add all monitored objects in the current sub-frequent item set to the first relationship graph, and connect the monitored objects pairwise to obtain the second relationship graph;

[0155] Traverse and determine whether all monitored objects in the associated sets in the relationship set exist in the second relationship graph;

[0156] If not, add all monitored objects in the current associated set to the second relationship graph, and connect the monitored objects pairwise to obtain an object relationship graph;

[0157] In the above determination process, if they exist, no operation is performed.

[0158] Obtain an event level weight, an object relationship weight, and an occurrence frequency weight based on the object relationship graph.

[0159] Preferably, the methods for obtaining the event level weight and the object relationship weight are:

[0160] All events in the object relationship graph correspond to corresponding level weights;

[0161] Traverse all events corresponding to each event object in the object relationship graph, and assign the highest level weight corresponding to all events to the corresponding event object as the event level weight .

[0162] Preferably, the number of connections corresponding to the monitored object with the largest number of connections in the object relationship graph is used as the maximum number of connections ;

[0163] Traverse all monitored objects in the object relationship graph, and based on the number of connections of the monitored object t and the maximum number of connections The ratio is used as the object relationship weight of the monitored object t :

[0164] .

[0165] Preferably, the method for obtaining the occurrence frequency weight is:

[0166] Traverse all objects in the object relationship graph in sequence, and count the number of times the object appears, including the number of times the event appears and the number of times the monitoring data appears ;

[0167] Obtain the maximum number of times the event appears and the maximum number of times the monitoring data appears that can be statistically obtained among all objects in the object relationship graph ;

[0168] Then the occurrence frequency weight of the object is:

[0169] .

[0170] The comprehensive score is obtained based on the event level weight, object relationship weight, and occurrence frequency weight.

[0171] Preferably, based on the event level weight , object relationship weight and occurrence frequency weight weighted to obtain the comprehensive score :

[0172] ;

[0173] wherein, , and represent the corresponding coefficients of the event level weight, object relationship weight, and occurrence frequency weight, which are customized according to actual needs.

[0174] Determine the key objects based on the comprehensive score.

[0175] Preferably, based on the comprehensive score is compared with the set threshold. If it is greater than the set threshold, it is determined as a key object.

[0176] Preferably, it further includes: obtaining the key area organization based on the key objects, specifically:

[0177] S901 Create a new organization R, and use the key objects not marked in the object relationship graph as the current key objects ;

[0178] S902 Select the objects in the object relationship graph that have a connection relationship with the current key object and the number of intermediate connection objects is less than the set value p and add them to the new organization, and mark the current key object ;

[0179] S903 Determine whether there are unmarked key objects based on the new organization ;

[0180] S904 If there are, select the objects in the object relationship graph that have a connection relationship with the unmarked key object and the number of intermediate connection objects is less than the set value p and add them to the new organization, and mark the unmarked key object ; If not, execute S906;

[0181] S905 Repeat S903 - S904 until all key objects in the new organization R are marked;

[0182] Repeat the execution of S901 - S905 until all the key objects in the object relationship graph are assigned to the corresponding organizations, and obtain multiple key area organizations.

[0183] Preferably, in actual situations, each organizational member has a direct or indirect relationship with the core member, i.e., the key object. The relationship intermediary in the indirect relationship generally does not exceed p. Taking the key object as the core, divide the organizational members to obtain the key area organizations.

[0184] Preferably, the set value p in this embodiment is less than 4.

[0185] Preferably, the present invention performs data fusion mining and analysis on the objects in border events and the recognized objects in monitoring data to form an associated relationship graph of border objects such as people, vehicles, ships, livestock, etc., and identifies key members based on the empowerment mode. The following technical problems are mainly solved:

[0186] 1) Based on border event information and border monitoring data, conduct event object relationship analysis, frequent item mining analysis of monitoring objects, relationship analysis of monitoring objects within the time - space range, etc., to form a border object relationship graph; propose an association analysis algorithm, whose idea is relatively simple to implement and takes into account the diversity of border objects; comprehensively consider border event data and monitoring data to make the obtained object relationship set more comprehensive.

[0187] 2) Creatively propose a method for identifying key members of a border object relationship graph based on event level and monitoring object frequency. The relationship graph makes the identification of key members interpretable and conducive to visualization, and there has been no relevant research before.

[0188] 3) Based on this, propose a method for dividing the organizational relationship of border objects with key members as the core, and provide an innovative algorithm for border object organization division.

[0189] Embodiment 3

[0190] As Figure 4 shown, a key object recognition system based on border object relationships includes: a first data processing module, a second data processing module, a third data processing module, a relationship graph generation module, a graph processing module, and a key object recognition module;

[0191] The first data processing module is used to obtain historical event information and obtain an associated event object set based on the historical event information;

[0192] The second data processing module is used to obtain monitoring data and conduct frequent item set mining to obtain the final frequent item set;

[0193] The third data processing module is used to obtain the monitored object set and its corresponding time and location based on the monitoring data; and obtain the relationship set of all events based on spatio-temporal rules based on the monitored object set and the time and location.

[0194] The relationship graph generation module is used to generate an object relationship graph based on the associated event object set, the final frequent item set, and the relationship set.

[0195] The graph processing module is used to obtain the event level weight, the object relationship weight, and the occurrence frequency weight based on the object relationship graph.

[0196] The key object recognition module is used to obtain a comprehensive score based on the event level weight, the object relationship weight, and the occurrence frequency weight; and determine the key objects based on the comprehensive score.

[0197] Preferably, in this embodiment, the functional implementation methods of each functional module correspond one by one to the above method content, and will not be elaborated here one by one.

[0198] Embodiment 4

[0199] Based on the same inventive concept, the present invention also provides a computer device, including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus.

[0200] The memory is used to store a computer program.

[0201] When the processor is used to execute the program stored in the memory, it can implement a key object recognition method based on the border object relationship as described in Embodiment 1 or 2.

[0202] The electronic device may include: a processor, a Communications Interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus. The processor can call the logical instructions in the memory to execute a key object recognition method based on the border object relationship as described in Embodiment 1 or 2.

[0203] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0204] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a key object recognition method and system based on border object relationships, having the following beneficial effects:

[0205] 1. The present invention proposes a key object recognition method for a border object relationship graph based on event level and monitoring object frequency. The relationship graph enables the key object recognition to be interpretable, is conducive to visual display, and thus improves the recognition efficiency of key objects.

[0206] 2. The algorithm of the present invention for calculating frequent items based on monitoring data and analyzing object relationships based on spatio-temporal rules can scientifically reflect the relationship set between objects, and thus can accurately identify key objects.

[0207] 3. The present invention comprehensively calculates the importance of an object by weighting factors such as event level, object connection, and occurrence frequency. The calculation result can reflect the importance of an object in the relationship graph, and organizes and divides the object set based on the key objects, and can efficiently and accurately obtain the border object organization.

[0208] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0209] The foregoing description of the disclosed embodiments enables those skilled in the art to practice or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A key object recognition method based on border object relationships, characterized in that Including: Obtain historical event information and obtain a set of associated event objects based on the historical event information; Obtain monitoring data and perform frequent item set mining to obtain the final frequent item set; Obtain a set of monitored objects and their corresponding time and location based on the monitoring data; Obtain a set of relationships of all events based on spatio-temporal rules based on the set of monitored objects and the time and location; Generate an object relationship graph based on the set of associated event objects, the final frequent item set, and the set of relationships; Obtain event level weights, object relationship weights, and occurrence frequency weights based on the object relationship graph; The methods for obtaining the event level weight and the object relationship weight are: All events in the object relationship graph correspond to corresponding level weights; Traverse all events corresponding to each event object in the object relationship graph, and assign the highest level weight corresponding to all events to the corresponding event object as the event level weight; Based on the number of connections corresponding to the monitored object with the largest number of connections in the object relationship graph as the maximum number of connections; Traverse all monitored objects in the object relationship graph, and use the ratio of the number of connections of the monitored object to the maximum number of connections as the object relationship weight; The method for obtaining the occurrence frequency weight is: Traverse all objects in the object relationship graph in sequence, and count the occurrence times of objects, including the occurrence times of events and the occurrence times of monitoring data Obtain the maximum number of occurrences of events that can be statistically obtained among all objects in the object relationship graph and the maximum number of occurrences of monitoring data Then the occurrence frequency weight of the object is as follows: Obtain a comprehensive score based on the event level weight, the object relationship weight, and the occurrence frequency weight; Determine key objects based on the comprehensive score.

2. The key object recognition method based on border object relationships according to claim 1, characterized in that The method for obtaining the set of associated event objects is: Obtain a border event object information association table including event identifiers and event object identifiers based on the historical event information; Statistically obtain a set of objects corresponding to multiple different events as the set of event objects based on the border event object information association table; Based on the association relationship of border events, merge the associated sets of event objects as the set of associated objects to obtain the set of associated event objects.

3. The key object recognition method based on border object relationships according to claim 2, characterized in that, The method for obtaining the final frequent item set is: Obtain a set of monitoring images of multiple frames based on the monitoring data; Obtain a set of monitored objects based on all monitored objects appearing in each frame of the monitoring image; Obtain multiple sub-frequent item sets based on the set of monitored objects; Take the union of all the sub-frequent item sets to obtain the final frequent item set.

4. The key object recognition method based on border object relationships according to claim 3, characterized in that Obtaining multiple sub-frequent item sets based on the set of monitored objects specifically includes: S201 Traverse each monitoring image in sequence based on the set of monitored objects; S202 Obtain the k-item set of the set of monitored objects based on the number of monitored objects in the monitoring image and the current number of frequent items k; S203 Obtain the sub-frequent item set corresponding to the number of frequent items k based on the support degree of each item in the k-item set; S204 Based on adding 1 to the number of frequent items k, repeat the above S201-S203 until no new frequent items are generated, and obtain multiple sub-frequent item sets corresponding to different numbers of frequent items.

5. The key object recognition method based on the border object relationship according to claim 4, characterized in that, The method for obtaining the k-item set is: Compare the number of monitored objects m in the monitoring image with the current number of frequent items k; If m > k, then obtain multiple k-item sets by permuting and combining k monitored objects in the current monitoring image; If m = k, then all monitored objects in the current monitored image are used as one of the k-item sets; If m < k, then the k-item set is not generated; Traverse all the monitored images, and all the obtained k-item sets form the k-item set.

6. The key object recognition method based on border object relationships according to claim 3, characterized in that, The method for obtaining the set of relationships of all events based on spatio-temporal rules is as follows: Traverse each monitored object in the monitored object set; Based on the current monitored object, successively judge the spatio-temporal coincidence with all other monitored objects in the monitored object set based on the time and location; When the spatio-temporal coincidence condition is met, there is an association relationship, and all association sets of the current monitored object are obtained; Otherwise, there is no association relationship; After traversing all the monitored objects, all the obtained association sets are used as the set of relationships of all events based on spatio-temporal rules.

7. The key object recognition method based on border object relationships according to claim 6, characterized in that The method for generating the object relationship graph is as follows: Create an initial relationship graph; Traverse and judge whether the event objects in all the associated object sets in the associated event object set exist in the initial relationship graph; If not, add all the event objects in the current associated object set to the initial relationship graph, and connect them pairwise based on the event objects to obtain the first relationship graph; Traverse and judge whether the monitored objects in all the sub-frequent item sets in the final frequent item set exist in the first relationship graph; If not, add all the monitored objects in the current sub-frequent item set to the first relationship graph, and connect them pairwise based on the monitored objects to obtain the second relationship graph; Traverse and judge whether the monitored objects in all the associated sets in the relationship set exist in the second relationship graph; If not, add all the monitored objects in the current associated set to the second relationship graph, and connect them pairwise based on the monitored objects to obtain the object relationship graph; In the above judgment process, if it exists, no operation is performed.

8. The key object recognition method based on the border object relationship according to claim 1, characterized in that, It further includes: Based on the key objects, key area organizations are obtained, specifically: S901 Create a new organization, and use the key objects not marked in the object relationship graph as the current key objects; S902 Select objects in the object relationship graph that have a connection relationship with the current key object and the number of intermediate connection objects is less than the set value and add them to the new organization, and mark the current key object; S903 Judge whether there are unmarked key objects based on the new organization; S904 If there are, select objects in the object relationship graph that have a connection relationship with the unmarked key object and the number of intermediate connection objects is less than the set value and add them to the new organization, and mark the unmarked key object; if not, execute S906; S905 Repeat S903 - S904 until all the key objects in the new organization are marked; S906 Repeat S901 - S905 until all the key objects in the object relationship graph are divided into corresponding organizations, and multiple key area organizations are obtained.

9. A key object recognition system based on border object relationships, which is used to execute a key object recognition method based on border object relationships according to any one of claims 1-8, characterized in that, It includes: The first data processing module, the second data processing module, the third data processing module, the relationship graph generation module, the graph processing module, and the key object recognition module; The first data processing module is used to obtain historical event information and obtain an associated event object set based on the historical event information; The second data processing module is used to obtain monitoring data and perform frequent item set mining to obtain a final frequent item set; The third data processing module is used to obtain a monitoring object set and its corresponding time and location based on the monitoring data; Obtain a relationship set of all events based on spatio-temporal rules based on the monitoring object set and the time and location; The relationship graph generation module is used to generate an object relationship graph based on the associated event object set, the final frequent item set, and the relationship set; The graph processing module is used to obtain an event level weight, an object relationship weight, and an occurrence frequency weight based on the object relationship graph; The key object recognition module is used to obtain a comprehensive score based on the event level weight, the object relationship weight, and the occurrence frequency weight; Determine key objects based on the comprehensive score.

Images