Repeated alarm optimization processing method and device, computer device and storage medium

Abstract

The application relates to a repeated alarm optimization processing method and device, computer equipment and a storage medium. In the case that it is determined that the duration of the historical track appearing in the current video frame collected by the camera meets the first preset condition of the unalarmed target person, whether a parking target having a space-time correlation relationship with the unalarmed target person exists is detected; if yes, the parking target having the space-time correlation relationship with the unalarmed target person is taken as a specified parking target; in the historical personnel having the space-time correlation relationship with the specified parking target, a first historical personnel appearing earlier than the unalarmed target person and having been subjected to alarm processing is determined; if the characteristic information of the unalarmed target person matches the characteristic information of the first historical personnel, it is determined that the ID of the unalarmed target person jumps, and alarm processing on the unalarmed target person is prohibited, thereby effectively avoiding the problem of repeated alarms on the same pedestrian caused by pedestrian ID jumping.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to a method, apparatus, computer equipment, and storage medium for optimizing the handling of recurring alarms. Background Technology

[0002] Urban expressways or highways are sections where pedestrians are prohibited. If pedestrians are seen on urban expressways or highways, it is usually because of a motor vehicle accident or breakdown. The people in the vehicle get out to check the vehicle condition or confirm the accident, etc. Road management personnel need to conduct a series of confirmation work for pedestrians on urban expressways or highways.

[0003] In existing technologies, cameras are installed on roads where pedestrians are generally not allowed, such as urban expressways or highways, to track and monitor pedestrians. Once a new pedestrian appears, an alarm is triggered and road management personnel are notified if the pedestrian meets certain conditions (such as persisting for a set time). However, the appearance of new pedestrians is often due to vehicle obstruction or changes in pedestrian posture causing pedestrian ID changes, leading to the problem of repeated alarms for the same pedestrian caused by these ID changes. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, and storage medium for optimizing the handling of repeated alarms, in order to solve the problem of repeated alarms for the same pedestrian caused by ID changes in related technologies.

[0005] In a first aspect, embodiments of this application provide a method for optimizing recurring alarm processing, the method comprising the following steps:

[0006] If a non-alarmed target person is identified in the current video frame captured by the camera whose duration of the historical trajectory meets the first preset condition, then among the recorded parking targets, it is detected whether there is a parking target that has a spatiotemporal correlation with the non-alarmed target person; the historical trajectory and the recorded parking targets are identified from the historical video frames.

[0007] If so, the parking target that has a spatiotemporal correlation with the target person who did not trigger an alarm will be designated as the parking target;

[0008] Among historical individuals who have a spatiotemporal correlation with the designated parking target, a first historical individual is identified whose appearance time is earlier than the target individual who did not trigger an alarm and who has already triggered an alarm; the historical individual is identified from the historical video frames.

[0009] If the feature information of the unalarmed target person matches the feature information of the first historical person, it is determined that the ID of the unalarmed target person has jumped, and the unalarmed target person is prohibited from being alarmed.

[0010] In some embodiments, the detecting whether there is a parking target having a spatiotemporal correlation with the unalarmed target person in the recorded parking target comprises the following steps:

[0011] In all recorded parking targets, it is detected whether there is a parking target having parking information satisfying a preset second condition, and if there is, the parking target satisfying the preset condition is determined as a parking target having a spatiotemporal correlation with the unalarmed target person; wherein the parking information is identified from the historical video frame;

[0012] The preset second condition comprises:

[0013] temporal and spatial.

[0014] In some embodiments, the detecting whether there is a parking target having parking information satisfying a preset second condition comprises the following steps:

[0015] Obtaining endpoint coordinate information of a detection frame of the parking target in a parking time period from the historical video frame;

[0016] Detecting whether the distance between the historical trajectory of the unalarmed target person and the detection frame of the parking target in the parking time period of the parking target is within a preset range.

[0017] In some embodiments, the method further comprises:

[0018] Detecting a motor vehicle target appearing in a video frame, and judging whether the motor vehicle target is in a parking state;

[0019] If the motor vehicle target is in a parking state, the motor vehicle target is taken as a parking target, and the duration, start time and end time of the parking state of the parking target are recorded.

[0020] In some embodiments, the method further comprises:

[0021] Saving the spatiotemporal correlation between the unalarmed target person and the specified parking target.

[0022] In some embodiments, the method further comprises:

[0023] If the feature information of the un-alarmed target person does not match the feature information of the first historical person or if there is no parking target having a spatio-temporal correlation with the un-alarmed target person, the un-alarmed target person is sent to an alarm process.

[0024] In some embodiments, the sending of the un-alarmed target person to the alarm process includes the following steps:

[0025] Determining whether the un-alarmed target person is a live target, and if the un-alarmed target person is a live target, performing alarm processing on the un-alarmed target person.

[0026] In some embodiments, the determining whether the un-alarmed target person is a live target includes the following steps:

[0027] Based on the spatio-temporal variation relationship and the motion feature of the un-alarmed target person, determining whether the un-alarmed target person is a live target; the spatio-temporal variation relationship and the motion feature of the un-alarmed target person are obtained from the historical video frames.

[0028] In a second aspect, the embodiments of the present application provide a repeated alarm optimization processing device, the device comprising: a detection module, a designation module, a determination module and a matching module;

[0029] The detection module is configured to, in a case where the duration of the historical trajectory appearing in the current video frame collected by the camera satisfies a first preset condition, detect, from the recorded parking targets, whether there is a parking target having a spatio-temporal correlation with the un-alarmed target person; the historical trajectory and the recorded parking target are identified from historical video frames.

[0030] The designation module is configured to, if so, designate the parking target having a spatio-temporal correlation with the un-alarmed target person as a designated parking target.

[0031] The determination module is configured to, from historical persons having a spatio-temporal correlation with the designated parking target, determine a first historical person appearing earlier than the un-alarmed target person and having been processed by alarm; the historical person is a person identified from the historical video frames.

[0032] The matching module is configured to, if the feature information of the un-alarmed target person matches the feature information of the first historical person, determine that the ID of the un-alarmed target person has jumped, and prohibit the un-alarmed target person from being processed by alarm.

[0033] In a third aspect, a computer device is provided in the present embodiment, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method of the first aspect when executing the computer program.

[0034] In a fourth aspect, a storage medium is provided in the present embodiment, having a computer program stored thereon, wherein the computer program implements the steps of the method of the first aspect when executed by a processor.

[0035] The repeated alarm optimization processing method and device, the computer device, and the storage medium described above, in the case where it is determined that the duration for which the historical trajectory appears in the current video frame collected by the camera satisfies the first preset condition and the target person is not alarmed, the recorded parking target is detected to determine whether there is a parking target that has a spatiotemporal correlation with the target person; the historical trajectory and the recorded parking target are identified from historical video frames; if so, the parking target that has a spatiotemporal correlation with the target person is taken as a specified parking target; in the historical personnel that have a spatiotemporal correlation with the specified parking target, a first historical personnel that appears earlier than the target person and has been subjected to alarm processing is determined; the historical personnel are identified from historical video frames; if the feature information of the target person matches the feature information of the first historical personnel, it is determined that the ID of the target person jumps, and the target person is prohibited from being subjected to alarm processing. The present application uses the spatiotemporal correlation between the target person and the parking target to perform personnel matching, determines whether the current target person is an already alarmed target person with ID jumping, and if it is determined that the current target person is an already alarmed target person with ID jumping, the current target person is not sent to the alarm process, effectively avoiding the problem of repeated alarms for the same pedestrian caused by pedestrian ID jumping. BRIEF DESCRIPTION OF DRAWINGS

[0036] The accompanying drawings, which are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and serve to explain the present application. In the drawings:

[0037] Figure 1 is an application scenario diagram of the repeated alarm optimization processing method according to an embodiment of the present application;

[0038] Figure 2 is a flowchart of the repeated alarm optimization processing method according to an embodiment of the present application;

[0039] Figure 3 is a structural schematic diagram of the repeated alarm optimization processing device according to an embodiment of the present application;

[0040] Figure 4 According to the structural schematic diagram of the computer device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is described and explained below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. Based on the embodiments provided by the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0042] Obviously, the drawings in the following description are only some examples or embodiments of the present application, and for those of ordinary skill in the art, the present application can be applied to other similar scenarios without creative labor on the basis of these drawings. In addition, it can be understood that although the efforts made in this development process can be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the present application, some design, manufacture or production changes on the basis of the technical content disclosed in the present application are only routine technical means and should not be understood as insufficient disclosure of the present application.

[0043] In the present application, the phrase "embodiment" means that the specific features, structures or characteristics described in combination with the embodiment can be contained in at least one embodiment of the present application. The appearance of this phrase at various places in the specification does not necessarily mean the same embodiment, nor is it an independent or alternative embodiment that is not mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in the present application can be combined with other embodiments without conflict.

[0044] Unless otherwise defined, technical terms and scientific terms used in the present application shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terms "a", "an", "one", "this", and the like, do not denote a quantity of particular aforementioned material, but rather denote that the names of the similar singular determiners "a", "an", "one", and "this" are used in the disclosure as a reference to one or more of the features so described. The terms "comprise", "comprising", "include", "including", "have", "has", "contain", "containing", or any other similar words, are intended to cover a non-exclusive inclusion. For example, a process, method, article, or apparatus that comprises a list of steps or components (units) is not limited to only those steps or components (units) that are explicitly listed, but can also include additional steps or components (units) that are not expressly listed, or can also include additional steps or components (units) that are inherent in the process, method, article, or apparatus. The terms "connect", "connected", "coupling", and the like, are not limited to direct or physical connections, but can include electrical connections, whether direct or indirect. The term "multiple" refers to two or more. The term "and / or" describes the associated relationship of associated objects, which means that there can be three relationships, for example, "A and / or B" can mean that A exists alone, A and B exist together, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects. The terms "first", "second", "third", and the like, are only used to distinguish similar objects, and do not represent a specific order of the objects.

[0045] Figure 1 The application scenario diagram of the repeated alarm optimization processing method provided for an embodiment of the present application is shown in FIG. 1. As shown in FIG. 1, the application scenario diagram of the repeated alarm optimization processing method provided for an embodiment of the present application includes the following steps. Figure 1As shown, data transmission can be performed between the server 101 and the terminal 102 through a network. The terminal 102 is configured to monitor the road conditions on a road on which pedestrians are generally not allowed to walk, such as an urban expressway or a highway, and collect monitoring data, and the server 101 is configured to analyze video frames collected by the terminal 102. In a case where a duration of a historical trajectory appearing in the video frames collected by the terminal 102 meets a first preset condition, the server 101 detects, from recorded parking targets, whether there is a parking target that has a spatiotemporal correlation with an unreported target person. The historical trajectory and the recorded parking target are identified from historical video frames. If so, the parking target that has the spatiotemporal correlation with the unreported target person is taken as a designated parking target. In historical persons, a first historical person that has a spatiotemporal correlation with the designated parking target and has been subjected to an alarm processing is determined. The historical person is identified from a video frame before a current video frame. If feature information of the target person matches feature information of the first historical person, it is determined that an ID of the unreported target person has jumped, and an alarm processing on the unreported target person is prohibited. The server 101 can be implemented by an independent server or a server cluster composed of multiple servers, and the terminal 102 can be implemented by an independent or multiple arbitrary video collection devices.

[0046] The embodiment provides a repeated alarm optimization processing method, as shown in the method, the method comprises the following steps: Figure 2

[0047] In a case where a duration of a historical trajectory appearing in the video frames collected by the terminal 102 meets a first preset condition, the server 101 detects, from recorded parking targets, whether there is a parking target that has a spatiotemporal correlation with an unreported target person. The historical trajectory and the recorded parking target are identified from historical video frames.

[0048] ​Specifically, due to the defects of the existing target detection algorithm, the personnel target may also be detected as a non-motor vehicle target sometimes, so all the unalarmed target personnel in the embodiment of the application can be referred to as unalarmed human body targets and unalarmed non-motor vehicle targets. Among them, the unalarmed target personnel can be the target personnel that have not been alarmed from the first appearance to now, or can be the target personnel whose ID jumps (after the ID jumps, the target personnel is not alarmed). In the embodiment of the application, the historical trajectory can be obtained by tracking detection on the historical video frames. The duration of the historical trajectory needs to meet the first preset condition, because the duration of the historical trajectory should not be too long or too short. If the time is too long, the target personnel has disappeared in the field of view of the camera before the current target personnel is processed. If the time is too short, the historical trajectory is not enough, and the spatio-temporal correlation between the target personnel and the parking target cannot be accurately obtained. Therefore, the duration of the historical trajectory in the application meets the first preset condition, which can be adjusted according to actual needs. In addition, the parking target in the embodiment of the application is a motor vehicle target in a parking state. Whether there is a parking target that has a spatio-temporal correlation with the unalarmed target personnel is determined, that is, whether there is a motor vehicle target in a parking time period that has a spatio-temporal correlation with the unalarmed target personnel.

[0049] Step S220, if yes, the parking target that has a spatio-temporal correlation with the unalarmed target personnel is taken as a specified parking target.

[0050] Specifically, assuming that the parking target A has a spatio-temporal correlation with the unalarmed target personnel m, the parking target A is taken as the specified parking target of the target personnel m.

[0051] Step S230, among the historical personnel that have a spatio-temporal correlation with the specified parking target, a first historical personnel that appears earlier than the unalarmed target personnel and has been processed is determined. The historical personnel is the personnel identified from the historical video frames.

[0052] Specifically, on a road where pedestrians are generally not allowed to walk, such as an urban expressway or a highway, a personnel target generally gets off a motor vehicle to perform a fault confirmation and the like, and then the ID can jump due to squatting up or a body shielding. Regardless of the ID before the jump or the ID after the jump, a spatiotemporal correlation relationship is generally generated with the same motor vehicle. Taking the above as an example, assuming that the specified parking target of the unalarmed target personnel m is A, it is indicated that the unalarmed target personnel m is most likely a historical personnel who has undergone ID jump and has a spatiotemporal correlation relationship with the specified parking target A. If the unalarmed target personnel m is indeed a personnel who has undergone ID jump, the personnel before the jump must have appeared earlier than the unalarmed target personnel m, otherwise there is no opportunity to become the unalarmed target personnel m after the jump. In addition, before the jump, there can be two situations. First, the ID before the jump is normally alarmed, and then there is no need to alarm the unalarmed target personnel m again. Second, the ID before the jump is missed, and then the unalarmed target personnel m needs to be alarmed. Therefore, in this embodiment, it is necessary to determine a first historical personnel who appears earlier than the unalarmed target personnel m and has been subjected to alarm processing from historical personnel who have a spatiotemporal correlation relationship with the specified parking target. If it is determined that the unalarmed target personnel m is a personnel who has undergone ID jump after the first historical personnel, there is no need to alarm the unalarmed target personnel m again. Assuming that the historical personnel who have a spatiotemporal correlation relationship with the specified parking target A are a, b, c, d, f, g, and h, only a, b, c, and d appear earlier than the unalarmed target personnel m and have been subjected to alarm processing, and then a, b, c, and d are determined as the first historical personnel.

[0053] In step S240, if the characteristic information of the unalarmed target personnel matches the characteristic information of the first historical personnel, it is determined that the ID of the unalarmed target personnel has jumped, and alarm processing on the unalarmed target personnel is prohibited.

[0054] Taking the above as an example, assuming that the designated parking target of the unalarmed target person m is A, and the first historical person a, b, c and d that have been processed for alarm exist in the spatiotemporal correlation relationship with A and appear earlier than the unalarmed target person m, the feature information of the unalarmed target person m is matched with the feature information of the first historical person a, b, c and d, so as to determine whether the unalarmed target person m is the alarmed person after ID jump. Specifically, the feature extraction algorithm can be used to extract the features of the first historical person a, b, c and d and the unalarmed target person m, so as to obtain the feature information of the unalarmed target person m and the feature information of the first historical person a, b, c and d. In addition, the existing cosine similarity calculation can be used to compare the similarity between the feature information of the unalarmed target person m and the feature information of the first historical person a, b, c and d. In addition, since the distance, posture and size of the captured person target in the video frame will affect the current extracted feature information, the video frame in the historical video frame of the first historical person a, b, c and d can be selected according to the distance, posture and size of the captured unalarmed target person m in the video frame, so as to improve the accuracy of the comparison result.

[0055] Assuming that the feature information of the unalarmed target person m matches the feature information of the first historical person a, it means that the unalarmed target person m is the target person after ID jump of the first historical person a. Since a is a person who has been processed for alarm, it is not necessary to alarm the unalarmed target person m again, so the alarm processing of the unalarmed target person m is prohibited, avoiding a series of operations caused by repeated alarm.

[0056] In the prior art, a camera is installed on a road such as a city expressway or a highway where pedestrians are generally not allowed to walk, and the camera is used to track and monitor pedestrians. Once a new pedestrian appears, as long as the new pedestrian meets the alarm condition (such as lasting for a set time), the new pedestrian will be processed for alarm and informed to the road supervision personnel. However, the appearance of the new pedestrian is most likely caused by the ID jump of the pedestrian due to the occlusion of the motor vehicle or the change of the posture of the pedestrian, which leads to the problem of repeated alarm of the same pedestrian caused by the ID jump of the pedestrian in the prior art.

[0057] To solve the above problems, the application provides a repeated alarm optimization processing method. In the case that the duration of the historical trajectory appearing in the current video frame collected by the camera meets the first preset condition, the method detects whether there is a parking target having a space-time correlation relationship with the unalarmed target personnel in the recorded parking target. The historical trajectory and the recorded parking target are identified from historical video frames. If so, the parking target having the space-time correlation relationship with the unalarmed target personnel is taken as a specified parking target. In the historical personnel having the space-time correlation relationship with the specified parking target, a first historical personnel appearing earlier than the unalarmed target personnel and having been processed for alarm is determined. The historical personnel is identified from the historical video frames. If the feature information of the unalarmed target personnel matches the feature information of the first historical personnel, it is determined that the ID of the unalarmed target personnel jumps, and the unalarmed target personnel is prohibited from being processed for alarm. The application utilizes the space-time correlation relationship between the unalarmed target personnel and the parking target to perform personnel matching, judges whether the current unalarmed target personnel is an ID-jumped and alarmed target, and if the current unalarmed target personnel is determined to be an ID-jumped and alarmed target, the current unalarmed target personnel is not sent to the alarm process, thereby effectively avoiding the problem of repeated alarm of the same pedestrian caused by the ID jump of the pedestrian.

[0058] As one of the embodiments, the step S210 detects whether there is a parking target having a space-time correlation relationship with the unalarmed target personnel in the recorded parking target, including the following steps:

[0059] Step S211, in all recorded parking targets, it is detected whether there is a parking target having parking information meeting a preset second condition, and if so, the parking target meeting the preset condition is determined as a parking target having a space-time correlation relationship with the unalarmed target personnel; wherein the parking information is identified from the historical video frames;

[0060] The preset second condition includes time and space.

[0061] In the embodiment, the parking information refers to the parking time and the parking location of the motor vehicle target, which can be identified from the historical video frames. By verifying the trajectory of the parking target and the trajectory of the unalarmed target person in time and space, it is determined whether the parking target is spatiotemporally related to the unalarmed target person. Specifically, assuming that the parking time period of the parking target B is 10:00-10:30, and the parking location of the parking target is S point, it is determined whether the trajectory of the unalarmed target person in 10:00-10:30 intersects with the parking target B, that is, whether the unalarmed target person appears at the S point or within a distance range from the S point determined according to the preset second condition in 10:00-10:30. Similarly, all the recorded parking targets are traversed to detect whether there is a parking target whose parking information satisfies the preset second condition, that is, whether there is a parking target whose trajectory in the parking time period is spatiotemporally intersected with the trajectory of the unalarmed target person.

[0062] Further, in one of the embodiments, the step S211 of detecting whether there is a parking target whose parking information satisfies the preset second condition comprises the following steps:

[0063] acquiring the endpoint coordinate information of the detection frame of the parking target in the parking time period from the historical video frames;

[0064] detecting whether the distance between the historical trajectory of the unalarmed target person and the detection frame of the parking target in the parking time period of the parking target is within a preset range.

[0065] Specifically, in the existing intelligent monitoring field, if the existence of the motor vehicle target is detected in the video frames, a detection frame is displayed around the motor vehicle target, the detection frame of the motor vehicle target is usually a rectangle and surrounds the motor vehicle target. In the embodiment, the endpoint coordinate information of the detection frame of the parking target in the parking time period is acquired from the historical video frames, the endpoint coordinate information of the detection frame reflects the position information of the parking target in the video frames, and then it is detected whether the distance between the historical trajectory of the unalarmed target person and the detection frame of the parking target in the parking time period of the parking target is within a preset range, thereby effectively detecting whether there is a parking target whose parking information satisfies the preset second condition.

[0066] In one of the embodiments, the repeated alarm optimization processing method further comprises the following steps:

[0067] In step S250, the motor vehicle target appearing in the video frames is detected to determine whether the motor vehicle target is in a parking state.

[0068] Specifically, the existing target detection algorithm can be used to detect the motor vehicle target appearing in the video frame, and determine whether the motor vehicle target is in the parking state, such as the public algorithm of SSD (Single Shot MultiBox Detector) or YOLO (You Only Look Once).

[0069] In step S260, if the current motor vehicle target is in the parking state, the current motor vehicle target is taken as the parking target, and the duration, the start time and the end time of the parking state of the current parking target are recorded.

[0070] If the current motor vehicle target is detected to be in the parking state, the current motor vehicle target can be taken as the parking target, and the duration, the start time and the end time of the parking state of the current parking target are recorded, so as to prepare for the subsequent spatiotemporal correlation between the target personnel.

[0071] As one of the embodiments, the spatiotemporal correlation between the target personnel without alarm and the designated parking target can be saved, the personnel matching efficiency is improved, and the working efficiency of the repeated alarm optimization processing of the present application is further improved.

[0072] In one of the embodiments, the repeated alarm optimization processing method further includes the following steps:

[0073] In step S270, if the feature information of the target personnel without alarm does not match the feature information of the first historical personnel or if there is no parking target having the spatiotemporal correlation with the target personnel without alarm, the target personnel without alarm is sent to the alarm process.

[0074] Specifically, if the feature information of the target personnel without alarm does not match the feature information of the first historical personnel or if there is no parking target having the spatiotemporal correlation with the target personnel without alarm, it is proved that the target personnel without alarm has not been processed by alarm, and then the target personnel without alarm is sent to the alarm process, so as to avoid the missed alarm of the target personnel without alarm, and further effectively avoid the impact on the road supervision work.

[0075] Further, in one of the embodiments, the step S270 of sending the target personnel without alarm to the alarm process includes the following steps:

[0076] In step S271, it is determined whether the target personnel without alarm is a living target, and in the case that the target personnel without alarm is a living target, the target personnel without alarm is processed by alarm.

[0077] Specifically, since the figurine on the motor vehicle and the figurine standing board, potted plant, columnar object and the like arranged on the road are likely to be detected as a personnel target, resulting in the false detection of the personnel target and false alarm, it is needed to determine whether the un-alarmed target personnel is a live target, and in the case that the un-alarmed target personnel is a live target, the un-alarmed target personnel is alarmed, so as to further effectively reduce the false alarm rate of the un-alarmed target personnel.

[0078] As one of the embodiments, whether the un-alarmed target personnel is a live target can be determined based on the space-time change relationship and the motion feature of the un-alarmed target personnel, wherein the space-time change relationship and the motion feature of the un-alarmed target personnel are obtained from the historical video frames. Whether the un-alarmed target personnel is a live target is determined based on the space-time change relationship and the motion feature of the un-alarmed target personnel, that is, whether the un-alarmed target personnel is a live target is determined by judging whether the duration of the continuous existence and the motion trajectory distance of the un-alarmed target personnel conform to the motion law of the live target, and judging whether the motion feature of the un-alarmed target personnel conforms to the motion feature of the live target.

[0079] Figure 3 is a schematic diagram of the repeated alarm optimization processing device according to the embodiment of the present application, as Figure 3 It is shown that a repeated alarm optimization processing device 30 is provided, which comprises a detection module 31, a designation module 32, a determination module 33 and a matching module 34.

[0080] The detection module 31 is configured to, in the case that the duration of the continuous existence of the historical trajectory in the current video frame collected by the camera satisfies the first preset condition, detect whether there is a parking target having a space-time correlation relationship with the un-alarmed target personnel in the recorded parking targets; the historical trajectory and the recorded parking target are identified from the historical video frames.

[0081] The designation module 32 is configured to, if yes, designate the parking target having the space-time correlation relationship with the un-alarmed target personnel as a designated parking target.

[0082] The determination module 33 is configured to determine a first historical personnel having an earlier appearance time than the un-alarmed target personnel and having been subjected to an alarm processing in the historical personnel having the space-time correlation relationship with the designated parking target; the historical personnel is a personnel identified from the historical video frames.

[0083] The matching module 34 is configured to, if the feature information of the target personnel matches the feature information of the first historical personnel, determine that the ID of the un-alarmed target personnel jumps, and prohibit the un-alarmed target personnel from being subjected to an alarm processing.

[0084] The repeated alarm optimization processing device 30 determines whether there is a parking target that has a spatiotemporal correlation with the unalarmed target person in the recorded parking targets in a case where the duration of the historical trajectory appearing in the current video frame collected by the camera meets the first preset condition; the historical trajectory and the recorded parking target are identified from historical video frames; if so, the parking target that has a spatiotemporal correlation with the unalarmed target person is taken as a designated parking target; a first historical person that appears earlier than the unalarmed target person and has been subjected to alarm processing is determined from historical persons that have a spatiotemporal correlation with the designated parking target; the historical person is identified from the historical video frames; if the feature information of the unalarmed target person matches the feature information of the first historical person, it is determined that the ID of the unalarmed target person has jumped, and the unalarmed target person is prohibited from being subjected to alarm processing. The application uses the spatiotemporal correlation between the unalarmed target person and the parking target to perform person matching, judges whether the current unalarmed target person is an ID-jumped already-alarm target, and if it is determined that the current unalarmed target person is an ID-jumped already-alarm target, the current unalarmed target person is not sent into the alarm process, effectively avoiding the problem of repeated alarms for the same pedestrian caused by pedestrian ID jumping.

[0085] In one of the embodiments, the detection module 31 is further configured to detect, from all the recorded parking targets, whether there is a parking target that has parking information meeting a preset second condition, and if so, determine the parking target meeting the preset condition as a parking target that has a spatiotemporal correlation with the unalarmed target person; wherein the parking information is identified from the historical video frames.

[0086] The preset second condition includes:

[0087] Spatiotemporal.

[0088] In one of the embodiments, the detection module 31 is further configured to acquire, from the historical video frames, endpoint coordinate information of the detection frame of the parking target in the parking time period.

[0089] Detect whether the distance between the historical trajectory of the unalarmed target person and the detection frame of the parking target in the parking time period of the parking target is within a preset range.

[0090] In one of the embodiments, the repeated alarm optimization processing device 30 further includes a recording module configured to detect a motor vehicle target appearing in a video frame, judge whether the motor vehicle target is in a parking state, take the current motor vehicle target as a parking target if the current motor vehicle target is in a parking state, and record the duration of the parking state, the parking start time, and the parking end time of the current parking target.

[0091] In one embodiment, the repeat alarm optimization processing device 30 further includes a storage module for storing the spatiotemporal correlation between non-alarm target personnel and designated parking targets.

[0092] In one embodiment, the repeated alarm optimization processing device 30 further includes an alarm module, which is used to send the non-alarmed target person into the alarm process if the characteristic information of the non-alarmed target person does not match the characteristic information of the first historical person or if there is no parking target with a spatiotemporal relationship with the non-alarmed target person.

[0093] In one embodiment, the alarm module is also used to determine whether the target personnel who have not been alarmed are living targets. If the target personnel who have not been alarmed are living targets, an alarm is triggered on the target personnel who have not been alarmed.

[0094] In one embodiment, the alarm module is further used to determine whether the non-alarmed target is a living target based on the spatiotemporal changes and motion characteristics of the non-alarmed target; the spatiotemporal changes and motion characteristics of the non-alarmed target are obtained from historical video frames.

[0095] It should be noted that the above modules can be functional modules or program modules, and can be implemented in software or hardware. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or they can be stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0096] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores a set of preset configuration information. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements the aforementioned optimized processing method for recurring alarms.

[0097] In one embodiment, a computer device is provided, which can be a terminal. The computer device comprises a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The network interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement a repeated alarm optimization processing method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0098] Those skilled in the art can understand that, Figure 4 The structure shown is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0099] In one embodiment, a storage medium is provided, which stores a computer program that is executed by a processor to implement the following steps:

[0100] In a case where it is determined that the duration of the occurrence of the unalarmed target person existing in the historical trajectory in the current video frame collected by the camera meets the first preset condition, in the recorded parking target, it is detected whether there is a parking target that has a spatiotemporal association relationship with the unalarmed target person; the historical trajectory and the recorded parking target are identified from historical video frames;

[0101] If yes, the parking target that has a spatiotemporal association relationship with the unalarmed target person is taken as a designated parking target;

[0102] Among the historical personnel that have a spatiotemporal association relationship with the designated parking target, a first historical personnel that appears earlier than the unalarmed target person and has been subjected to alarm processing is determined; the historical personnel is a personnel identified from historical video frames;

[0103] If the feature information of the target person matches the feature information of the first historical personnel, it is determined that the ID of the unalarmed target person has jumped, and the unalarmed target person is prohibited from being subjected to alarm processing.

[0104] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0105] Among all the recorded parking targets, it is detected whether there is a parking target whose parking information satisfies a preset second condition, and if there is, the parking target satisfying the preset condition is determined as a parking target having a spatio-temporal correlation with the target person not reported.

[0106] The preset second condition includes:

[0107] In time and in space.

[0108] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0109] The endpoint coordinate information of the detection frame of the parking target in the parking time period is obtained from the historical video frames.

[0110] It is detected whether the distance between the historical trajectory of the target person not reported and the detection frame of the parking target in the parking time period of the parking target is within a preset range.

[0111] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0112] The motor vehicle target appearing in the video frame is detected, and it is judged whether the motor vehicle target is in a parking state.

[0113] If the current motor vehicle target is in a parking state, the current motor vehicle target is taken as a parking target, and the duration of the parking state of the current parking target, the parking start time and the parking end time are recorded.

[0114] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0115] The spatio-temporal correlation relationship between the target person not reported and the specified parking target is saved.

[0116] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0117] If the feature information of the target person not reported does not match the feature information of the first historical person or if there is no parking target having a spatio-temporal correlation with the target person not reported, the target person not reported is sent to an alarm process.

[0118] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0119] determine whether the unalarmed target person is a living target, and perform an alarm process on the unalarmed target person in a case where the unalarmed target person is a living target.

[0120] In one embodiment, the processor, when executing the computer program, also implements the following steps:

[0121] determine whether the unalarmed target person is a living target based on the spatiotemporal variation relationship and the motion feature of the unalarmed target person; and the spatiotemporal variation relationship and the motion feature of the unalarmed target person are obtained from the historical video frames.

[0122] In the above storage medium, in a case where the duration of the historical trajectory appearing in the current video frame collected by the camera satisfies the first preset condition, the unalarmed target person is detected from the recorded parking targets to determine whether there is a parking target having a spatiotemporal correlation relationship with the unalarmed target person; the historical trajectory and the recorded parking target are identified from the historical video frames; if so, the parking target having the spatiotemporal correlation relationship with the unalarmed target person is taken as a specified parking target; a first historical person appearing earlier than the unalarmed target person and having been processed for an alarm is determined from historical persons having the spatiotemporal correlation relationship with the specified parking target; the historical person is identified from the historical video frames; if the feature information of the unalarmed target person matches the feature information of the first historical person, it is determined that the ID of the unalarmed target person has jumped, and the unalarmed target person is prohibited from being processed for an alarm. The application utilizes the spatiotemporal correlation relationship between the unalarmed target person and the parking target to perform personnel matching, determines whether the current unalarmed target person is an ID-jumped and already-alarmed target person, and if it is determined that the current unalarmed target person is an ID-jumped and already-alarmed target person, the current unalarmed target person is not sent into an alarm process, effectively avoiding the problem of repeated alarm on the same pedestrian caused by the ID jumping of the pedestrian.

[0123] It should be understood that the specific embodiments described herein are merely exemplary and are not intended to limit the application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments provided by the application without creative labor are within the scope of protection of the application.

[0124] Obviously, the drawings are only some examples or embodiments of the application, and can be applied to other similar cases without creative labor according to the drawings for those of ordinary skill in the art. In addition, it can be understood that although the work done in the development process may be complex and long, certain design, manufacture or production changes made by those of ordinary skill in the art according to the disclosed technical content of the application are only routine technical means and should not be regarded as insufficient disclosure of the application.

[0125] The word "implementation" in this application refers to the specific features, structures, or characteristics described in connection with an implementation can be included in at least one implementation of the present application. The phrase appears in various places in the specification does not necessarily mean the same implementation, nor does it mean independence or alternatives available to each other. It is clear or implicitly understood by those skilled in the art that the embodiments described in this application can be combined without conflict with other embodiments.

[0126] The above embodiments only express several implementation manners of the present application, which are described in detail and specifically, but cannot be understood as the limitation of the patent protection scope. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present application, which all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for optimizing the handling of recurring alarms, characterized in that, The method comprises the following steps: In the case that the duration of the historical trajectory appearing in the current video frame collected by the camera satisfies the first preset condition, detecting whether there is a parking target having a spatio-temporal correlation with the unreported target person in the recorded parking target; the historical trajectory and the recorded parking target are identified from historical video frames; If yes, taking the parking target having a spatio-temporal correlation with the unreported target person as a designated parking target; In the historical personnel having a spatio-temporal correlation with the designated parking target, determining a first historical personnel appearing earlier than the unreported target person and having been processed for alarm; the historical personnel is identified from the historical video frames; If the feature information of the unreported target person matches the feature information of the first historical personnel, it is determined that the ID of the unreported target person has jumped, and the unreported target person is prohibited from being processed for alarm; If the feature information of the unreported target person does not match the feature information of the first historical personnel or if there is no parking target having a spatio-temporal correlation with the unreported target person, the unreported target person is sent into an alarm process.

2. The repeated alarm optimization process of claim 1, wherein, The method further comprises the following steps: In all recorded parking targets, detecting whether there is a parking target having parking information satisfying a preset second condition, and if yes, determining the parking target satisfying the preset condition as a parking target having a spatio-temporal correlation with the unreported target person; wherein the parking information is identified from the historical video frames; The preset second condition comprises: temporal and spatial.

3. The repeated alarm optimization process of claim 2, wherein, The method further comprises the following steps: Obtaining endpoint coordinate information of a detection frame of the parking target in a parking time period from the historical video frames; Detecting whether the distance between the historical trajectory of the unreported target person and the detection frame of the parking target in the parking time period of the parking target is within a preset range.

4. The repeated alarm optimization process of claim 1, wherein, The method further comprises: detecting a motor vehicle target appearing in a video frame, and judging whether the motor vehicle target is in a parking state; If the motor vehicle target is in a parking state, taking the motor vehicle target as a parking target, and recording the duration of the parking state, the parking start time and the parking end time of the parking target.

5. The repeated alarm optimization process of claim 1, wherein, The method further comprises: saving the spatio-temporal correlation between the unreported target person and the designated parking target.

6. The repeated alarm optimization process of claim 1, wherein, The method further comprises the following steps: judging whether the unreported target person is a living target, and processing the unreported target person for alarm in the case that the unreported target person is a living target.

7. The repeated alarm optimization process of claim 6, wherein, The method further comprises the following steps: The method comprises: determining whether the non-alarming target person is a living target based on a spatiotemporal variation relationship and a motion feature of the non-alarming target person; and obtaining the spatiotemporal variation relationship and the motion feature of the non-alarming target person from the historical video frames.

8. A repeated alarm optimization processing apparatus characterized by comprising: The device comprises a detection module, a designation module, a determination module, a matching module, and an alarm module. The detection module is configured to, in a case where a duration in which the historical trajectory appears in the current video frame collected by the camera satisfies a first preset condition, detect, from the recorded parking targets, whether there is a parking target that has a spatiotemporal correlation relationship with the non-alarming target person; the historical trajectory and the recorded parking target are identified from historical video frames. The designation module is configured to, if so, designate the parking target that has the spatiotemporal correlation relationship with the non-alarming target person as a designated parking target. The determination module is configured to, from historical persons that have the spatiotemporal correlation relationship with the designated parking target, determine a first historical person that appears earlier than the non-alarming target person and has been subjected to an alarm process; the historical person is identified from the historical video frames. The matching module is configured to, if feature information of the non-alarming target person matches feature information of the first historical person, determine that an ID of the non-alarming target person has jumped, and prohibit the non-alarming target person from being subjected to the alarm process. The alarm module is configured to, if the feature information of the non-alarming target person does not match the feature information of the first historical person or if there is no parking target that has the spatiotemporal correlation relationship with the non-alarming target person, send the non-alarming target person to an alarm process.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the steps of the method in any one of claims 1 to 7.

10. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 7.

Images