CN116071818A - Personal injury behavior detection method, device, equipment and storage medium - Google Patents
Personal injury behavior detection method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN116071818A CN116071818A CN202211624985.6A CN202211624985A CN116071818A CN 116071818 A CN116071818 A CN 116071818A CN 202211624985 A CN202211624985 A CN 202211624985A CN 116071818 A CN116071818 A CN 116071818A
- Authority
- CN
- China
- Prior art keywords
- behavior
- real
- action
- determining
- video stream
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000006378 damage Effects 0.000 title claims abstract description 44
- 208000027418 Wounds and injury Diseases 0.000 title claims abstract description 43
- 208000014674 injury Diseases 0.000 title claims abstract description 43
- 238000001514 detection method Methods 0.000 title claims abstract description 38
- 230000006399 behavior Effects 0.000 claims abstract description 163
- 238000004458 analytical method Methods 0.000 claims abstract description 103
- 230000000875 corresponding effect Effects 0.000 claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 14
- 238000012216 screening Methods 0.000 claims description 10
- 241000607479 Yersinia pestis Species 0.000 claims description 8
- 238000009432 framing Methods 0.000 claims description 3
- 206010001488 Aggression Diseases 0.000 claims description 2
- 230000016571 aggressive behavior Effects 0.000 claims description 2
- 208000012761 aggressive behavior Diseases 0.000 claims description 2
- 230000036544 posture Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 239000006071 cream Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000000981 bystander Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012106 screening analysis Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000002459 sustained effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Human Computer Interaction (AREA)
- Social Psychology (AREA)
- Computational Linguistics (AREA)
- Psychiatry (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Image Analysis (AREA)
Abstract
The application discloses a method, a device, equipment and a storage medium for detecting personal injury behaviors, and relates to the technical field of safety detection. The method comprises the following steps: acquiring a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set; based on the analysis target set and the action behaviors, determining a protection object with defensive action behaviors in the analysis target set; and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion. Through the technical scheme of the application, the recognition accuracy of personal injury behavior detection can be effectively improved, and behavior scenes such as campus spoofing are effectively aimed at.
Description
Technical Field
The present invention relates to the field of security detection technologies, and in particular, to a method, an apparatus, a device, and a storage medium for detecting personal injury behaviors.
Background
The campus safety problem relates to the stability of society, and more promotes thousands of families. Campus spoofing events occur daily in various countries of the world and there are sometimes reports of campus student spoofing violent cases, some of which are rather severe in nature. The frequent occurrence of campus deception events affects campus safety, disturbs the normal order of schools, and has very bad influence on students. Any form of spoofing is unacceptable, and the spoofing causes harm to "victims" and "bystanders" in addition to "victims". Therefore, schools, society and the like should pay attention to the phenomenon of campus spoofing, develop prevention and cure in time, and prevent the phenomenon of spoofing from spreading further.
Disclosure of Invention
Accordingly, the present invention is directed to a method, apparatus, device and storage medium for detecting personal injury, which can effectively identify the scenario of illegal infringement in progress and effectively aim at the scenario of campus spoofing. The specific scheme is as follows:
in a first aspect, the present application discloses a method for detecting personal injury behavior, comprising:
acquiring a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set;
based on the analysis target set and the action behaviors, determining a protection object with defensive action behaviors in the analysis target set;
and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion.
Optionally, the acquiring a real-time video stream and performing video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and a corresponding action behavior of a pedestrian target in the analysis target set, including:
acquiring a real-time video stream, and framing the real-time video stream to obtain different video frames to be detected;
identifying different pedestrian targets in the video frame to be detected by using a preset target detection method, and marking the pedestrian targets to obtain an analysis target set of the pedestrian targets with corresponding marks;
and acquiring corresponding action behaviors of pedestrian targets in the analysis target set by using a first preset human body posture estimation algorithm.
Optionally, the determining, based on the analysis target set and the action behavior, a protection object having defensive action behavior in the analysis target set includes:
screening pedestrian targets of which the residence time in the analysis target set exceeds a first preset duration and the duration time of the defensive action behavior exceeds a second preset duration, so as to obtain targets to be judged;
and judging the target to be judged by using a preset judging rule, and determining the target to be judged meeting the preset judging rule as the protection object.
Optionally, the determining the target to be determined by using a preset determination rule, and determining the target to be determined that meets the preset determination rule as the protection object includes:
stacking the behavior judgment results of the targets to be judged, and judging whether the judgment success ratio of the behavior judgment results is larger than a preset proportion threshold value or not within a third preset duration;
and when the judging success ratio of the behavior judging result is larger than a preset proportion threshold value, determining the target to be judged as the protection object.
Optionally, the method for detecting personal injury behavior further includes:
and acquiring the historical behavior data of the protected object, and determining the time corresponding to the first piece of the historical behavior data in the historical behavior data as the event starting time of the illegal invasive behavior scene.
Optionally, the determining the target area picture of the protection object on the real-time video stream and determining the suspected harmful person with the aggressive action in the target area picture includes:
determining a target area picture of the protection object on the real-time video stream, and clustering pedestrian targets around the protection object in the target area picture by using a density clustering algorithm to obtain a corresponding clustering result;
and acquiring corresponding action behaviors of the pedestrian targets in the clustering result by using a second preset human body posture estimation algorithm, and screening harmful suspects with aggressive action behaviors.
Optionally, after determining the target area picture of the protection object on the real-time video stream and determining the offensive suspects with aggressive actions in the target area picture to identify the current illegal infringement action scene according to the protection object and the offensive suspects, the method further includes:
performing face recognition on the protected object, and comparing a face recognition result with a face database constructed in advance to confirm the identity information of the protected object;
generating corresponding alarm information according to the identity information, and sending the alarm information to a control system so as to carry out alarm stopping on the behavior action of the suspected harmful person through the control system.
In a second aspect, the present application discloses a personal injury behavior detection device comprising:
the video analysis module is used for acquiring a real-time video stream and carrying out video analysis on the real-time video stream so as to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set;
the protection object determining module is used for determining a protection object with defensive action behaviors in the analysis target set based on the analysis target set and the action behaviors;
the pest suspected person determining module is used for determining a target area picture of the protection object on the real-time video stream and determining a pest suspected person with aggressive action in the target area picture so as to identify the current illegal attack action scene according to the protection object and the pest suspected person.
In a third aspect, the present application discloses an electronic device comprising a processor and a memory; wherein the memory is for storing a computer program that is loaded and executed by the processor to implement the personal injury behavior detection method as described above.
In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program; wherein the computer program when executed by a processor implements a personal injury behavior detection method as described above.
The application provides a personal injury behavior detection method, which comprises the steps of obtaining a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set; based on the analysis target set and the action behaviors, determining a protection object with defensive action behaviors in the analysis target set; and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion. As can be seen, since there are often similar action patterns in attack behavior and many actions, sports behavior, playing alarm, etc., the action patterns are not fixed and are numerous, and are often misreported in practical application. Therefore, in the application, by analyzing the real-time video stream, an analysis target set is determined, and after the action behaviors of the analysis targets in the analysis target set are obtained, the protection object with defensive action behaviors is determined, that is, the defensive actions are detected first. Then detecting the suspected harmful person with aggressive action based on the target area picture around the protected object, and judging the aggressive action of the suspected harmful person. Therefore, from the fixed and small quantity of defensive actions, the characteristics that the action modes of defensive action are fixed and the quantity is small are utilized, the scene of illegal infringement actions which are happening is effectively identified, the more efficient and more accurate detection of the human infringement actions is realized, the false alarm rate is reduced, the calculation efficiency is improved, and the method is effective for scenes such as campus cream.
In addition, the personal injury behavior detection device, the personal injury behavior detection equipment and the storage medium provided by the application correspond to the personal injury behavior detection method, and the effects are the same as those of the personal injury behavior detection method.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for detecting personal injury behavior disclosed in the present application;
FIG. 2 is a flowchart of a specific method for detecting personal injury behavior disclosed in the present application;
FIG. 3 is a schematic diagram of a personal injury behavior detection device disclosed in the present application;
fig. 4 is a block diagram of an electronic device disclosed in the present application.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
At present, the frequency of campus spoofing events affects campus safety, disturbs the normal order of schools, and has very bad influence on students. Therefore, the personal injury behavior detection scheme can effectively identify the scenes of illegal infringement behaviors in the process of happening, and effectively aims at scenes such as campus spoofing.
The embodiment of the invention discloses a method for detecting personal injury behaviors, which is shown in fig. 1 and comprises the following steps:
step S11: and acquiring a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set.
In the embodiment of the application, the camera is used for detecting personal safety injury behaviors, the real-time video stream is obtained through the camera, and then video analysis is carried out on the real-time video stream. Specifically, firstly framing the real-time video stream to obtain different video frames to be detected; and then, identifying different pedestrian targets in the video frame to be detected by using a preset target detection method, and marking the pedestrian targets to obtain an analysis target set of the pedestrian targets with corresponding marks.
It can be understood that the same or different pedestrian targets exist in different video frames to be detected, the pedestrian targets are identified by using a preset target detection method and are correspondingly marked, and the same target can be identified in a plurality of continuous video frames. For example, a Yolo algorithm may be adopted to analyze and identify a target in a video image to be detected, and then a unique Identification ID is given to the same target in different video frames to be detected by using a ReID (Person Re-Identification technology), so that the obtained analysis target set includes different pedestrian targets and corresponding marks.
Further, a first preset human body posture estimation algorithm is utilized to obtain corresponding action behaviors of pedestrian targets in the analysis target set. Because the scene in the real-time video stream is analyzed to judge whether illegal infringement behavior exists in the current scene, the behavior action of each pedestrian target in the analysis target set is acquired by using a preset human body posture estimation algorithm. In a recommended scheme, a gesture action evaluation method for identifying skeletal actions such as PoseC3D can be used for extracting a human skeleton sequence and identifying gesture actions of targets to obtain an identification result of the action actions of the pedestrian targets in the analysis target set.
Step S12: and determining a protection object with defensive action behaviors in the analysis target set based on the analysis target set and the action behaviors.
In the embodiment of the application, according to the action behaviors of different pedestrian targets in the analysis target set, the protection object is identified, namely, the target currently in illegal attack is determined. Such targets are often accompanied by defensive action behaviors including various self-protecting behaviors of the human body when attacked, including but not limited to, tightening, head-protecting, and the like.
It can be understood that in a campus scenario, students often form groups of three and five, because the sports and playing alarm, many limb actions and attack actions have similarity, so the action modes of the attack actions are not fixed and the number is large, if the aggressive behavior actions are judged, misjudgment is easy to generate, and more false recognition situations exist in actual use. According to the embodiment of the application, the problem is avoided, the pedestrian target with the defensive action behavior in the analysis target set is determined to serve as the protection object from the fixed and small defensive action behavior cut-in, the recognition accuracy is effectively improved, and the landing practicability is higher.
In the embodiment of the application, the video frame to be detected where the protection object is located is traced back, and the event starting time of the current illegal invasive behavior scene is confirmed. Specifically, the historical behavior data of the protection object is obtained, and the time corresponding to the first piece of historical behavior data in the historical behavior data is determined as the event starting time of the illegal infringement behavior scene.
In this embodiment of the present application, as a target to be analyzed in an analysis target set that is not determined as a protection target, if defensive action behaviors occur, the target is stored in historical behavior data of the target. When the object to be analyzed is determined to be a protected object, the time of the first piece of data in the historical behavior data is searched, and the time is determined to be the event starting time of the illegal infringement behavior scene. It should be noted that, here, the historical behavior data stores the result of successful judgment of the defensive action of the pedestrian target in the analysis target set and the current judgment time, and the unsuccessful judgment is not recorded. In addition, in order to prevent the situation that the historical behavior data occupy too much memory or the judgment result is wrong because the time interval between the latest defensive action behavior and the first defensive action behavior is too long, an automatic expiration time of the historical behavior data can be preset, for example, 5 minutes are preset, and the historical data exceeding 5 minutes are deleted, so that the identification accuracy and the detection performance can be further improved, and the event starting time can be more accurately determined.
Step S13: and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion.
In the embodiment of the application, after the protection object is determined, the suspected pest is required to be locked. That is, by analyzing the pedestrian targets around the protected object, the target of the suspected person who is being harmed illegally is determined.
Specifically, determining a target area picture of the protection object on the real-time video stream, and clustering pedestrian targets around the protection object in the target area picture by using a density clustering algorithm to obtain a corresponding clustering result; and acquiring corresponding action behaviors of the pedestrian targets in the clustering result by using a second preset human body posture estimation algorithm, and screening harmful suspects with aggressive action behaviors.
In the embodiment of the application, clustering analysis is performed on pedestrian targets around a protection object of a picture in a real-time video stream. It can be understood that in the situation that illegal infringement behavior exists, bystanders or partnerships of suspected harmful people can exist, and the density clustering algorithm is utilized to perform cluster analysis on pedestrian targets around the protected object to obtain corresponding clustering results. Then, the positions and the postures of pedestrians in different clustering results are obtained through a preset human posture estimation algorithm, targets with aggressive action behaviors are screened out through pedestrian posture judgment, the targets are used as harmful suspects, and the targets can be put into a first suspects set; and for the clustering result which is not put into the first suspects set but suspected to be partnered with a person in the first suspects set, a second suspects set can be put into the clustering result, so that the partnering of the analysis and the locking of the first suspects are realized.
The application provides a personal injury behavior detection method, which comprises the steps of obtaining a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set; based on the analysis target set and the action behaviors, determining a protection object with defensive action behaviors in the analysis target set; and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion. As can be seen, since there are often similar action patterns in attack behavior and many actions, sports behavior, playing alarm, etc., the action patterns are not fixed and are numerous, and are often misreported in practical application. Therefore, in the application, by analyzing the real-time video stream, an analysis target set is determined, and after the action behaviors of the analysis targets in the analysis target set are obtained, the protection object with defensive action behaviors is determined, that is, the defensive actions are detected first. Then detecting the suspected harmful person with aggressive action based on the target area picture around the protected object, and judging the aggressive action of the suspected harmful person. Therefore, from the fixed and small quantity of defensive actions, the characteristics that the action modes of defensive action are fixed and the quantity is small are utilized, the scene of illegal infringement actions which are happening is effectively identified, the more efficient and more accurate detection of the human infringement actions is realized, the false alarm rate is reduced, the calculation efficiency is improved, and the method is effective for scenes such as campus cream.
The embodiment of the application discloses a specific personal injury behavior detection method, which is shown in fig. 2 and comprises the following steps:
step S21: and acquiring a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set.
For more specific processing in step S21, reference may be made to the corresponding content disclosed in the foregoing embodiment, and no further description is given here.
Step S22: and screening pedestrian targets of which the residence time exceeds a first preset duration and the duration time of the defensive action behavior exceeds a second preset duration in the analysis target set so as to obtain targets to be judged.
In the embodiment of the application, each pedestrian target in the analysis target set in the real-time video stream is analyzed, the action behaviors of the pedestrian targets are judged, and the protection object with defensive action behaviors is determined. In a specific implementation process, judging the walking state of the pedestrian targets in the analysis target set, ignoring non-stationary targets, and screening stationary targets. And screening out the pedestrian targets with the stay time exceeding the first preset duration in the screening analysis target set as stationary targets. For example, 10 seconds may be preset as the preset duration of the stay determination, and if the stay time of the pedestrian target exceeds 10 seconds, it is selected.
Further, in order to prevent erroneous judgment due to occasional actions, a guard action duration judgment time may be preset, and judgment that the defensive action occurs for a time longer than a second preset duration is determined as that the target is actually defending. Therefore, the embodiment of the application judges whether the object is a protection object or not by screening the object which is static and has defensive actions and taking the object as the object to be judged.
Step S23: and judging the target to be judged by using a preset judging rule, and determining the target to be judged meeting the preset judging rule as the protection object.
In the embodiment of the application, whether the target to be judged is a protection object is judged through a preset judging rule. Specifically, stacking the behavior judgment results of the target to be judged, and judging whether the judgment success ratio of the behavior judgment results is larger than a preset proportion threshold value or not within a third preset duration; and when the judging success ratio of the behavior judging result is larger than a preset proportion threshold value, determining the target to be judged as the protection object. That is, a behavior determination result stack within a third preset time period is set for each target to be determined. For example, a behavior determination result stack of the last 10 seconds is set for each target to be determined, and if the behavior determination result is larger than a preset proportion threshold value within 10 seconds, the target is set as a protection object. The number of times of judgment of the defensive action behavior, which is the judgment of the power ratio, divided by the total number of times of judgment is generally determined on a frame-by-frame basis or a frame skip basis.
Step S24: and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion.
For more specific processing in step S24, reference may be made to the corresponding content disclosed in the foregoing embodiment, and no further description is given here.
Step S25: and carrying out face recognition on the protected object, and comparing the face recognition result with a face database constructed in advance to confirm the identity information of the protected object.
In the embodiment of the application, after the protection object and the suspected harmful person are determined, the current illegal infringement behavior scene can be alarmed. The identity information of the protected object, namely the victim, is confirmed by comparing the face recognition of the protected object with the existing face library, such as a campus resident face library, a community personnel face library and the like, so that corresponding alarm information is generated according to the identity information of the protected object to carry out alarm.
Step S26: generating corresponding alarm information according to the identity information, and sending the alarm information to a control system so as to carry out alarm stopping on the behavior action of the suspected harmful person through the control system.
In the embodiment of the application, the generated alarm information is sent to a control system, such as a command center, a security department, a public security organization and the like, according to the identity information of the protected object. And the current behaviors of the suspected pest are restrained on site through the control system. For example, after the injury behavior is identified, the injury behavior is prevented from being sustained by quickly playing pre-recorded audio through on-site sound playing equipment to warn the suspected person; remote speaking can also be performed through the control system to alarm and stop the behavior actions of the suspected person.
The application provides a personal injury behavior detection method, which comprises the steps of obtaining a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set; screening pedestrian targets of which the residence time in the analysis target set exceeds a first preset duration and the duration time of the defensive action behavior exceeds a second preset duration, so as to obtain targets to be judged; judging the target to be judged by using a preset judging rule, and determining the target to be judged meeting the preset judging rule as the protection object; determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with an aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion; performing face recognition on the protected object, and comparing a face recognition result with a face database constructed in advance to confirm the identity information of the protected object; generating corresponding alarm information according to the identity information, and sending the alarm information to a control system so as to carry out alarm stopping on the behavior action of the suspected harmful person through the control system. As can be seen, since there are often similar action patterns in attack behavior and many actions, sports behavior, playing alarm, etc., the action patterns are not fixed and are numerous, and are often misreported in practical application. Therefore, in the application, by analyzing the real-time video stream, an analysis target set is determined, and after the action behaviors of the analysis targets in the analysis target set are obtained, the protection object with defensive action behaviors is determined, that is, the defensive actions are detected first. Then detecting the suspected harmful person with aggressive action based on the target area picture around the protected object, and judging the aggressive action of the suspected harmful person. Therefore, from the fixed and small quantity of defensive actions, the characteristics that the action modes of defensive action are fixed and the quantity is small are utilized, the scene of illegal infringement actions which are happening is effectively identified, the more efficient and more accurate detection of the human infringement actions is realized, the false alarm rate is reduced, the calculation efficiency is improved, and the method is effective for scenes such as campus cream.
Correspondingly, the embodiment of the application also discloses a personal injury behavior detection device, as shown in fig. 3, which comprises:
the video analysis module 11 is configured to obtain a real-time video stream, and perform video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set;
a protection object determining module 12, configured to determine, based on the analysis target set and the action behavior, a protection object having defensive action behavior in the analysis target set;
the suspected harmful person determining module 13 is configured to determine a target area picture of the protected object on the real-time video stream, and determine a suspected harmful person having an aggressive action in the target area picture, so as to identify a current illegal attack action scene according to the protected object and the suspected harmful person.
The more specific working process of each module may refer to the corresponding content disclosed in the foregoing embodiment, and will not be described herein.
Therefore, through the scheme of the embodiment, a real-time video stream is obtained, and video analysis is performed on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set; based on the analysis target set and the action behaviors, determining a protection object with defensive action behaviors in the analysis target set; and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion. As can be seen, since there are often similar action patterns in attack behavior and many actions, sports behavior, playing alarm, etc., the action patterns are not fixed and are numerous, and are often misreported in practical application. Therefore, in the application, by analyzing the real-time video stream, an analysis target set is determined, and after the action behaviors of the analysis targets in the analysis target set are obtained, the protection object with defensive action behaviors is determined, that is, the defensive actions are detected first. Then detecting the suspected harmful person with aggressive action based on the target area picture around the protected object, and judging the aggressive action of the suspected harmful person. Therefore, from the fixed and small quantity of defensive actions, the characteristics that the action modes of defensive action are fixed and the quantity is small are utilized, the scene of illegal infringement actions which are happening is effectively identified, the more efficient and more accurate detection of the human infringement actions is realized, the false alarm rate is reduced, the calculation efficiency is improved, and the method is effective for scenes such as campus cream.
Further, the embodiment of the present application further discloses an electronic device, and fig. 4 is a block diagram of an electronic device 20 according to an exemplary embodiment, where the content of the figure is not to be considered as any limitation on the scope of use of the present application.
Fig. 4 is a schematic structural diagram of an electronic device 20 according to an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input output interface 25, and a communication bus 26. The memory 22 is used for storing a computer program, and the computer program is loaded and executed by the processor 21 to implement relevant steps in the personal injury behavior detection method disclosed in any of the foregoing embodiments. In addition, the electronic device 20 in the present embodiment may be a computer.
In this embodiment, the power supply 23 is configured to provide an operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and an external device, and the communication protocol to be followed is any communication protocol applicable to the technical solution of the present application, which is not specifically limited herein; the input/output interface 25 is used for acquiring external input data or outputting external output data, and the specific interface type thereof may be selected according to the specific application requirement, which is not limited herein.
The memory 22 may be a carrier for storing resources, such as a read-only memory, a random access memory, a magnetic disk, or an optical disk, and the resources stored thereon may include an operating system 221, a computer program 222, data 223, and the like, and the data 223 may include various data. The storage means may be a temporary storage or a permanent storage.
The operating system 221 is used for managing and controlling various hardware devices on the electronic device 20 and computer programs 222, which may be Windows Server, netware, unix, linux, etc. The computer program 222 may further include a computer program that can be used to perform other specific tasks in addition to the computer program that can be used to perform the personal injury behavior detection method performed by the electronic device 20 disclosed in any of the foregoing embodiments.
Further, embodiments of the present application disclose a computer readable storage medium, where the computer readable storage medium includes random access Memory (Random Access Memory, RAM), memory, read-Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, magnetic disk, or optical disk, or any other form of storage medium known in the art. The computer program, when executed by the processor, implements the personal injury behavior detection method. For specific steps of the method, reference may be made to the corresponding contents disclosed in the foregoing embodiments, and no further description is given here.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, so that the same or similar parts between the embodiments are 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 points refer to the description of the method section.
The steps of a personal injury behavior detection or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above description of the method, the device, the equipment and the storage medium for detecting personal injury provided by the invention is provided in detail, and specific examples are applied to the description of the principle and the implementation mode of the invention, and the description of the above examples is only used for helping to understand the method and the core idea of the invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.
Claims (10)
1. A method for detecting personal injury behavior, comprising:
acquiring a real-time video stream, and carrying out video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set;
based on the analysis target set and the action behaviors, determining a protection object with defensive action behaviors in the analysis target set;
and determining a target area picture of the protection object on the real-time video stream, and determining a harmful suspicion with aggressive action in the target area picture so as to identify a current illegal infringement action scene according to the protection object and the harmful suspicion.
2. The method for detecting personal injury behavior according to claim 1, wherein the acquiring a real-time video stream and performing video analysis on the real-time video stream to determine an analysis target set in the real-time video stream and a corresponding action behavior of a pedestrian target in the analysis target set comprises:
acquiring a real-time video stream, and framing the real-time video stream to obtain different video frames to be detected;
identifying different pedestrian targets in the video frame to be detected by using a preset target detection method, and marking the pedestrian targets to obtain an analysis target set of the pedestrian targets with corresponding marks;
and acquiring corresponding action behaviors of pedestrian targets in the analysis target set by using a first preset human body posture estimation algorithm.
3. The method for detecting personal injury behavior according to claim 1, wherein the determining a protection object having defensive action behavior in the analysis target set based on the analysis target set and the action behavior comprises:
screening pedestrian targets of which the residence time in the analysis target set exceeds a first preset duration and the duration time of the defensive action behavior exceeds a second preset duration, so as to obtain targets to be judged;
and judging the target to be judged by using a preset judging rule, and determining the target to be judged meeting the preset judging rule as the protection object.
4. The personal injury behavior detection method according to claim 3, wherein the determining the target to be determined using a preset determination rule and determining the target to be determined that satisfies the preset determination rule as the protection target includes:
stacking the behavior judgment results of the targets to be judged, and judging whether the judgment success ratio of the behavior judgment results is larger than a preset proportion threshold value or not within a third preset duration;
and when the judging success ratio of the behavior judging result is larger than a preset proportion threshold value, determining the target to be judged as the protection object.
5. The personal injury behavior detection method according to claim 1, further comprising:
and acquiring the historical behavior data of the protected object, and determining the time corresponding to the first piece of the historical behavior data in the historical behavior data as the event starting time of the illegal invasive behavior scene.
6. The method for detecting personal injury behavior according to claim 1, wherein the determining a target area picture of the protection object on the real-time video stream and determining a suspected person having an aggressive action in the target area picture comprises:
determining a target area picture of the protection object on the real-time video stream, and clustering pedestrian targets around the protection object in the target area picture by using a density clustering algorithm to obtain a corresponding clustering result;
and acquiring corresponding action behaviors of the pedestrian targets in the clustering result by using a second preset human body posture estimation algorithm, and screening harmful suspects with aggressive action behaviors.
7. The personal injury behavior detection method according to any one of claims 1 to 6, wherein after determining a target area picture of the protection object on the real-time video stream and determining a offensive suspect having an aggressive behavior in the target area picture to identify a current illegal offensive behavior scene from the protection object and the offensive suspect, further comprising:
performing face recognition on the protected object, and comparing a face recognition result with a face database constructed in advance to confirm the identity information of the protected object;
generating corresponding alarm information according to the identity information, and sending the alarm information to a control system so as to carry out alarm stopping on the behavior action of the suspected harmful person through the control system.
8. A personal injury behavior detection device, comprising:
the video analysis module is used for acquiring a real-time video stream and carrying out video analysis on the real-time video stream so as to determine an analysis target set in the real-time video stream and corresponding action behaviors of pedestrian targets in the analysis target set;
the protection object determining module is used for determining a protection object with defensive action behaviors in the analysis target set based on the analysis target set and the action behaviors;
the pest suspected person determining module is used for determining a target area picture of the protection object on the real-time video stream and determining a pest suspected person with aggressive action in the target area picture so as to identify the current illegal attack action scene according to the protection object and the pest suspected person.
9. An electronic device, characterized in that, the electronic device includes a processor and a memory; wherein the memory is for storing a computer program that is loaded and executed by the processor to implement the personal injury behavior detection method of any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program; wherein the computer program when executed by a processor implements a personal injury behavior detection method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211624985.6A CN116071818A (en) | 2022-12-16 | 2022-12-16 | Personal injury behavior detection method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211624985.6A CN116071818A (en) | 2022-12-16 | 2022-12-16 | Personal injury behavior detection method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116071818A true CN116071818A (en) | 2023-05-05 |
Family
ID=86177864
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211624985.6A Pending CN116071818A (en) | 2022-12-16 | 2022-12-16 | Personal injury behavior detection method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116071818A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118116084A (en) * | 2024-04-22 | 2024-05-31 | 浙江深象智能科技有限公司 | Behavior information detection method, apparatus and storage medium |
-
2022
- 2022-12-16 CN CN202211624985.6A patent/CN116071818A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118116084A (en) * | 2024-04-22 | 2024-05-31 | 浙江深象智能科技有限公司 | Behavior information detection method, apparatus and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4924607B2 (en) | Suspicious behavior detection apparatus and method, program, and recording medium | |
| CN109598885B (en) | Monitoring system and alarm method thereof | |
| CN107590439B (en) | Target person identification and tracking method and device based on surveillance video | |
| CN110263633B (en) | Method, system and storage medium for detecting and early warning of toxic person based on space-time correlation | |
| CN110705477A (en) | Behavior analysis method and apparatus, electronic device, and computer storage medium | |
| CN109033973B (en) | Monitoring and early warning method and device and electronic equipment | |
| US20240095862A1 (en) | Method for determining dangerousness of person, apparatus, system and storage medium | |
| CN110738178A (en) | Garden construction safety detection method and device, computer equipment and storage medium | |
| CN107886667A (en) | Alarm method and device | |
| EP2779133A2 (en) | System and method of anomaly detection | |
| US8494159B2 (en) | System and practice for surveillance privacy-protection certification and registration | |
| CN113330491B (en) | Electronic gate opening method and device and server | |
| CN111985428A (en) | Security detection method and device, electronic equipment and storage medium | |
| CN109615256A (en) | Wisdom garden security protection risk control method, storage medium, electronic equipment and system | |
| CN112288975A (en) | Event early warning method and device | |
| CN114937237A (en) | Construction site safety monitoring method and system based on AI intelligent identification | |
| CN110717357B (en) | Early warning method and device, electronic equipment and storage medium | |
| CN116071818A (en) | Personal injury behavior detection method, device, equipment and storage medium | |
| CN110689694B (en) | Intelligent monitoring system and method based on image processing | |
| US10878245B2 (en) | Video content activity regions | |
| Flammini et al. | Petri net modelling of physical vulnerability | |
| US20210034689A1 (en) | Method and apparatus for detecting regional event based on search engine, and storage medium | |
| CN112511807A (en) | Fire fighting access monitoring linkage method | |
| CN117354469A (en) | District monitoring video target tracking method and system based on security precaution | |
| Kaluža et al. | Detection of suspicious behavior from a sparse set of multiagent interactions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |