CN113486777A

CN113486777A - Behavior analysis method and device for target object, electronic equipment and storage medium

Info

Publication number: CN113486777A
Application number: CN202110748884.9A
Authority: CN
Inventors: 陈兆斌; 刘利; 杨立东; 郎桦; 张静普
Original assignee: Beijing Eway Dacheng Technology Co ltd
Current assignee: Beijing Eway Dacheng Technology Co ltd
Priority date: 2021-07-02
Filing date: 2021-07-02
Publication date: 2021-10-08

Abstract

The embodiment of the specification discloses a behavior analysis method and device for a target object, electronic equipment and a storage medium, wherein the target object entering a target area is monitored through a millimeter wave radar to obtain target track information; configuring a camera according to the target track information so as to capture a dynamic target object by using the camera to obtain continuous frame pictures; drawing a target object posture picture through the continuous frame pictures; and continuously and dynamically recognizing, analyzing and predicting the motion of the target object attitude picture so as to analyze the real behavior of the target object.

Description

Behavior analysis method and device for target object, electronic equipment and storage medium

Technical Field

The embodiment of the specification relates to the technical field of security and protection, in particular to a behavior analysis method and device for a target object, electronic equipment and a storage medium.

Background

The means for realizing perimeter precaution in the market at present are mainly divided into three categories:

1) electronic perimeter security alarm system

The electronic perimeter alarm system comprises infrared correlation, vibration cables, leakage cables, electronic fences and other modes, and prevention is carried out in a physical enclosure mode;

however, these security measures can only be installed on the wall, on the ground, etc., and are affected by the surrounding environment, weather, and surrounding electromagnetic interference, and the identification performance is poor, the false alarm rate is high, and the recording and tracing capabilities are lacking.

2) Intelligent video monitoring system

The intelligent video monitoring system processes, analyzes and understands video signals by utilizing a computer vision technology, positions, identifies and tracks changes in a monitored scene by automatically analyzing sequence images without human intervention, analyzes and judges the behavior of a target on the basis, can send out an alarm or provide useful information in time when abnormal conditions occur, effectively assists safety personnel in handling crisis, and reduces false alarm and missing alarm phenomena to the maximum extent;

however, in an actual environment, illumination change, complexity of target motion, shielding, similarity of colors of a target and a background, a cluttered background and the like all increase the difficulty of target detection and tracking algorithm design, and if the target is in a foggy and snowy weather, any identification cannot be performed due to picture reasons, so that the detection accuracy is still a main problem under some special conditions;

3) radar dome camera linked system

The principle of the radar dome camera linkage system is that a millimeter wave radar detects a target by transmitting and receiving high-frequency electromagnetic waves, a rear-end signal processing module calculates target information such as existence, speed, direction, distance, angle and the like of a moving object by using echo signals, the target information is processed by an upper computer program and then is sent to a monitoring camera system, and the real-time tracking of the target is realized by setting PTZ parameters of a camera, and the target is reported to the monitoring system to be checked and processed by related personnel through real-time video;

the alarm device has the advantages that the alarm rate is low, and the influence of light and shadow change, rain, snow, haze, dust, background change and shielding is avoided; the radar dome camera linkage system has the advantages that although the system can accurately detect information such as target coordinates, speed and the like, real behaviors of targets cannot be identified, when farmlands, roads and other activity places exist outside perimeter guardrails, human activities are more, and safety risks are relatively high.

Disclosure of Invention

In view of this, embodiments of the present disclosure provide a method and an apparatus for analyzing behavior of a target object, an electronic device, and a storage medium, so as to solve the problem that the false alarm rate caused by normal activities of people or vehicles in a monitored area cannot be effectively reduced in the prior art.

The embodiment of the specification adopts the following technical scheme:

an embodiment of the present specification provides a method for analyzing a behavior of a target object, including:

monitoring a target object entering a target area through a millimeter wave radar to obtain target track information;

configuring a camera according to the target track information so as to capture a dynamic target object by using the camera to obtain continuous frame pictures;

drawing a target object posture picture through the continuous frame pictures;

and continuously and dynamically recognizing, analyzing and predicting the motion of the target object attitude picture so as to analyze the real behavior of the target object.

An embodiment of the present specification further provides a device for analyzing behavior of a target object, including:

the acquisition module monitors a target object entering a target area through a millimeter wave radar to acquire target track information;

the configuration module is used for configuring the camera according to the target track information so as to capture a dynamic target object by using the camera to obtain continuous frame pictures;

the picture drawing module is used for drawing a target object posture picture through the continuous frame pictures;

and the analysis module is used for carrying out continuous dynamic recognition analysis and action prediction on the target object posture picture so as to analyze the real behavior of the target object.

An embodiment of the present specification further provides an electronic device for analyzing target object behaviors, which includes at least one processor and a memory, where the memory stores a program and is configured to enable the at least one processor to execute the following steps:

drawing a target object posture picture through the continuous frame pictures;

Embodiments of the present specification also provide a computer readable storage medium having computer readable instructions stored thereon, the computer readable instructions being executable by a processor to implement behavioral analysis of a target object.

The embodiment of the specification adopts at least one technical scheme which can achieve the following beneficial effects:

when the image is shielded and blurred, the operations of target monitoring, track drawing, alarming and the like can be continuously carried out through the detection means of the millimeter wave radar, the usability of the system can not be influenced, and therefore the problem that the intelligent video monitoring system cannot realize monitoring work under the condition that the video image is blurred is solved.

By adding a human body posture recognition algorithm and a machine learning algorithm, the system is increased in behavior analysis capability, the behavior and the basic intention of a target appearing in a camera image can be analyzed in real time, whether the suspicious target is a normal behavior or a dangerous behavior is determined, and when the suspicious target is classified as the normal behavior, the system only prompts no alarm; when the classification is dangerous behavior, alarming and related alarming measures are carried out according to a preset strategy, meanwhile, model optimization can be continuously carried out through a machine learning algorithm, and the accuracy of behavior analysis algorithm processing is continuously improved, so that the problem that the alarm rate is high due to the fact that human or vehicle normal activities exist in a detection area of a linkage system of the radar dome camera is solved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the embodiments of the specification and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the specification and together with the description serve to explain the application and not to limit the application. In the drawings:

fig. 1 is a schematic flowchart of a method for analyzing behavior of a target object according to an embodiment of the present disclosure;

fig. 2 is a schematic flowchart of a method for analyzing behavior of a target object according to an embodiment of the present disclosure;

FIG. 3 is an interaction diagram of the millimeter-wave radar, the upper computer and the camera in FIG. 2;

FIG. 4 is a diagram of a specific human body posture picture;

FIG. 5 is a body posture picture of a human leg-raising action;

FIG. 6 is a diagram of a simulated arming scheme for an airport perimeter;

fig. 7 is a schematic workflow diagram of a target object behavior analysis system according to an embodiment of the present disclosure;

FIG. 8 is a general defense diagram constructed by the airport behavior analysis system according to the above-mentioned target object;

fig. 9 is a schematic structural diagram of a target object behavior analysis apparatus provided in an embodiment of the present specification.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more clear, the technical solutions of the present application will be clearly and completely described below with reference to the specific embodiments of the present specification and the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present disclosure, and not all embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step are within the scope of the present application.

The technical solutions provided by the embodiments of the present description are described in detail below with reference to the accompanying drawings.

Fig. 1 is a schematic flow chart of a method for analyzing a behavior of a target object according to an embodiment of the present disclosure.

S101: and monitoring a target object entering a target area through a millimeter wave radar to acquire target track information.

In the embodiment of the present specification, the millimeter wave radar may be a radar whose operating frequency band is in the millimeter wave frequency band, and detects a target by transmitting and receiving a high-frequency electromagnetic wave, and the back-end signal processing module calculates target information such as existence, speed, direction, distance, angle, and the like of a moving object by using an echo signal.

The target area comprises a defense area, an early warning area and an alarm area.

In a specific application scenario, the target track information includes a target coordinate, a target speed, a target angle, and a target distance.

S103: and configuring a camera according to the target track information so as to capture a dynamic target object by using the camera to obtain continuous frame pictures.

As an application example, configuring a camera according to the target track information may include:

and configuring the PTZ value of the camera according to the target track information.

Wherein, the PTZ value of the camera is P (pan), which represents the movement of the camera in the horizontal direction, namely the rotation parameter of the camera; t (tilt), which represents the movement of the camera in the vertical direction, i.e. the elevation parameter of the camera lens; z (zoom), which represents the zooming of the camera, i.e. adjusting the focal length parameters of the camera.

Specifically, when the camera captures a dynamic target object, the target object is within a preset distance range of a center of a monitoring image of the camera.

Therefore, related monitoring personnel can check and record the target object and the surrounding environment of the target object more clearly and accurately.

Wherein, the camera is a 360-degree rotating high-definition night vision dome camera.

S105: and drawing a target object posture picture through the continuous frame pictures.

As an application embodiment, the drawing of the target object pose picture by the continuous frame pictures may include:

performing algorithm processing on the continuous frame pictures by adopting an object posture recognition algorithm;

and obtaining the target object posture picture according to the algorithm processing result.

Further, performing algorithm processing on the continuous frame pictures by using an object posture recognition algorithm may include:

extracting feature maps of the continuous frame pictures through a convolutional neural network;

marking and connecting key points of the skeleton of the target object according to the characteristic diagram to obtain a point line diagram of the whole skeleton of the target object;

obtaining the target object posture picture according to the algorithm processing result, wherein the method comprises the following steps:

and drawing the target object posture picture according to the point line graph of the whole framework of the target object.

S107: and continuously and dynamically recognizing, analyzing and predicting the motion of the target object attitude picture so as to analyze the real behavior of the target object.

As an application embodiment, performing continuous dynamic recognition analysis and motion prediction on the target object posture picture to analyze the real behavior of the target object may include:

continuously inputting the obtained target object posture pictures into a machine learning algorithm;

extracting key feature points from the continuously input target object attitude pictures by adopting a convolutional neural network;

calculating a core feature vector of the key feature points;

and observing the position change of the key characteristic points within a preset time range by using the core characteristic vector so as to analyze the real behavior of the target object according to the position change.

The convolutional neural network is a feature model obtained through deep learning of a plurality of human body posture pictures, and the feature model comprises human body bones, human face expressions, actions and posture feature models.

In particular, the key feature points may include joints having degrees of freedom on the human body, and the joints may include neck, shoulder, elbow, wrist, waist, knee, ankle.

Further, the method may further include:

and judging the real behavior of the target object to determine whether the real behavior is dangerous behavior.

After determining the real behavior of the target object, the method may further include:

if the real behavior is judged to be normal behavior and the target object is located in the defense area and/or the early warning area, prompting the target object to avoid the target object from entering the warning area by mistake;

and if the real behavior is judged to be dangerous behavior and the target object is located in the early warning area and/or the warning area, warning according to a preset strategy.

According to the behavior analysis method for the target object, when the image is shielded and blurred, the operations of target monitoring, trajectory drawing, alarming and the like can be continuously performed through the detection means of the millimeter wave radar, the usability of the system cannot be affected, and therefore the problem that the intelligent video monitoring system cannot achieve monitoring work under the condition that the video image is blurred is solved.

As shown in fig. 2, a flow diagram of a method for analyzing a behavior of a target object according to an embodiment of the present specification is provided.

S201, monitoring a target human body entering a target area through a millimeter wave radar, tracking and positioning the target human body, predicting a target track according to a built-in algorithm, predicting a plurality of coordinate points in the future, wherein the number of predicted coordinate points is generally 3-5, and obtaining target track information;

s202, the millimeter wave radar uploads the acquired target track information to an upper computer, and the upper computer obtains the PTZ value for driving a camera to track a target human body after acquiring the target track information and calculating through a series of algorithms;

s203, setting the PTZ value of the camera to enable the camera to capture a dynamic target in real time to obtain continuous frame pictures;

s204, the camera returns the obtained continuous frame pictures to the upper computer, and the upper computer performs algorithm processing on the frame pictures to draw human body posture pictures;

s205, carrying out continuous dynamic recognition analysis and motion prediction on the human body posture picture to analyze the real behavior of the target human body;

s206, when the real implementation is classified as normal behavior, only prompting and not alarming; and when the real behavior is classified into dangerous behavior, alarming according to a preset strategy and taking a preset alarming measure.

Fig. 3 is a schematic diagram illustrating interaction among the millimeter-wave radar, the upper computer, and the camera in the above application embodiment.

Specifically, the step S203 may include that the upper computer performs algorithm processing on the frame picture by using a human body posture recognition algorithm, extracts the feature map in the frame picture through a convolutional neural network, performs key point marking and connection on human bones according to the feature map, obtains a point line graph of the human body overall skeleton, and finally draws the human body posture picture, as shown in fig. 4, the schematic diagram of the specific human body posture picture is shown.

Step S204 may include continuously inputting the obtained human body posture picture into a machine learning algorithm, which extracts key points from the continuously input human body posture picture by using a convolutional neural network technique, calculates core feature vectors of the key points, adds a time sequence, and observes position changes of the key points within a preset time range to analyze real behaviors of the human body according to a position change result, as shown in fig. 5, which is a human body posture picture of a leg raising motion of the human body.

Fig. 6 shows a diagram of a simulated defense scheme for the perimeter of an airport.

And dividing the real scene into a defense area, an early warning area and an alarm area, wherein the area division is performed through a virtual warning line drawn by the system.

The defense deploying area is used for detecting and is a buffer area of the early warning area, and personnel and vehicles enter the defense deploying area to display the target position and track, but do not carry out target identification;

the early warning area is a key monitoring area, personnel and vehicles enter the early warning area, the system enters an early warning state, the system starts the video monitoring system to automatically track, record and identify the target, the acousto-optic and electric warning system can be selectively started according to a warning scheme set by the system, the system analyzes, identifies and marks the target through the radar three-dimensional coordinate data, and displays real-time track and position information;

the alarm area is the most serious area, when a target (personnel, vehicles and animals) enters the alarm area or approaches the alarm area, the height change (barrier crossing behavior) system automatically alarms, the video image system is started to automatically track, record and identify the target, the acousto-optic-electric alarm system is started according to the warning scheme set by the system, and the system analyzes, identifies and marks the target through the three-dimensional coordinate data of the radar and displays real-time track and position information.

Fig. 7 is a schematic workflow diagram of the target object behavior analysis system.

When suspicious people illegally enter the perimeter warning area, after the radar detects the suspicious people, animals, vehicles and the like, the radar processes the coordinates, angles, distances, advancing tracks and the like of the target and predicts the tracks through a pre-judgment algorithm, meanwhile, the actual target information and the pre-judgment information are sent to an upper computer of a control center, the upper computer obtains the real position and the pre-judgment position of the target after algorithm processing, and sets parameters such as a PTZ value of the camera, the camera is driven to realize video monitoring linkage, the camera is quickly positioned to an alarm area to track the target in real time, image information generated in the tracking process is sent to a behavior analysis subsystem in real time to analyze the target behavior, when the image is unclear, the focal length of the camera is automatically adjusted to adjust the monitoring video, if the situation that the image cannot be identified due to weather or shielding and the like occurs, target detection and tracking are automatically carried out by the millimeter wave radar, so that the condition that monitoring is invalid due to the fact that the camera loses the target is avoided;

when the image is clear, the target is automatically modeled, analyzed and marked, whether the target behavior is a preset dangerous action (the situations of fence shearing, fence crossing, airport inherent facility damage, guardrail collision, illegal invasion, crowd gathering and the like) is judged, when the target behavior is judged not to be the dangerous action, the information is automatically filtered and recorded for subsequent tracing, when the target behavior is judged to be the dangerous action, an application system is timely notified to give an alarm, and after a series of calculations, the application system finally displays information such as the nature of the alarm, the place of the case, the time, a site plan and the like on a screen, and related monitoring personnel are reminded of timely and effective processing.

After the behavior analysis module is additionally arranged, a brain is equivalently added to the radar dome camera contact system, and various behaviors of the target in the image are intelligently analyzed, so that the system can predict the behavior intention of the target at the initial stage of finding the target, the alarm condition caused by most normal behaviors is eliminated, the usability and the safety of the system are greatly improved, and meanwhile, the intelligent analysis system enables machines to share more work, so that the identification accuracy is ensured, and the workload and the labor intensity of staff are greatly reduced.

At present, in an actual scene, the following dangerous behaviors are preset, and the dangerous behaviors mainly comprise: collision enclosure, crossing of the guardrail, throwing action, gathering of people, overtime detention, described in detail in the following table:

after the system is on line, the system can really realize the goals of all-weather, high performance, high efficiency and safety all day long, and can ensure the safety of the perimeter of the airport to the maximum extent.

At present, according to the test results, the following objectives have been achieved:

fig. 8 shows an overall defense diagram constructed by the airport according to the target object behavior analysis system.

As shown in fig. 8, a circle of radar dome camera is arranged inside the airport perimeter near the guardrail, the millimeter wave radar is responsible for target detection inside and outside the perimeter, and the camera is normally monitored.

When the millimeter wave radar detects a suspicious target, whether the suspicious target reaches a distance threshold is judged according to a system strategy, and when the suspicious target reaches the distance threshold, information such as the position, the speed and the angle of the target is sent to an upper computer of a control center for information processing, and then real-time target tracking is carried out by setting a PTZ value of a camera, so that the target is ensured to be always inside a video-accessible region.

The camera sends the real-time video key frame to a behavior analysis system to identify and analyze the behavior of the target, determines whether dangerous behaviors (conditions of guardrail net shearing, guardrail net crossing, airport inherent facility damage, guardrail collision, illegal invasion, crowd gathering and the like) exist in the target, immediately reports the dangerous behaviors to an application system to give an alarm and is delivered to monitoring personnel to process in time if the dangerous behaviors exist, and records and archives real-time information to ensure that the later period is documented.

Its advantages mainly include:

high accuracy: the behavior analysis is carried out aiming at radar detection imaging and video monitoring images, the continuity analysis of the whole process of posture-action-behavior is achieved through the processing and analysis of continuous images, the current suspicious target state (people, animals and vehicles, throwing objects at high altitude, crossing fences, damaging fences and the like, or illegally breaking vehicles, or causing the intrusion of animals which are harmful to airport operation and the like) is obtained, and in addition, the continuous learning of machines according to different behaviors at the later stage is added, so that the very high alarm accuracy is guaranteed;

low false alarm rate: the safety protection technology of multiple means (millimeter wave radar, camera monitoring, human body posture recognition, machine learning) and hierarchy (millimeter wave radar target tracking and thermal imaging, linkage video monitoring and behavior analysis) is adopted, the millimeter wave radar is used for tracking the suspicious target in real time by utilizing the flexibility linkage video monitoring, the position information and the image information of the suspicious target are guaranteed to be updated by the system at any time, various preset behaviors are analyzed by multiple artificial intelligence means such as machine learning, and the like, so that reliable information guarantee is provided for security personnel to handle abnormal conditions, the safety of a target area is guaranteed, the accuracy of alarming is also guaranteed, and the false alarm rate is reduced;

timely processing: the millimeter wave radar and the monitoring video near the millimeter wave radar and the monitoring video can be linked in time to find and track a suspicious target in a large-scale real-time manner through the radar, the behavior analysis can quickly perform behavior analysis on the suspicious target and make judgment, the alarm system immediately alarms to inform an attendant of the suspicious target, the radar detection and the video monitoring of the system always keep tracking the suspicious target before abnormal conditions are processed, and the position information of the suspicious target is provided for security personnel or service personnel in real time, so that the attendant can quickly and accurately process the abnormal conditions;

the system can also carry out continuous relearning according to the field disposal records, and a knowledge base is accumulated for technical defense work while the false alarm rate is reduced.

Wherein the behavior analysis device of the target object may include:

an obtaining module 901, which monitors a target object entering a target area through a millimeter wave radar to obtain target track information;

a configuration module 902, configured to configure a camera according to the target track information, so as to capture a dynamic target object by using the camera, and obtain continuous frame pictures;

a picture drawing module 903 for drawing the target object posture picture through the continuous frame pictures;

and the analysis module 904 is used for performing continuous dynamic recognition analysis and motion prediction on the target object posture picture so as to analyze the real behavior of the target object.

Further, the method further comprises:

Further, the target area comprises a defense area, an early warning area and an alarm area;

after the real behavior of the target object is judged, the method further comprises the following steps:

Further, the target track information includes target coordinates, target speed, target angle, and target distance.

Further, configuring a camera according to the target track information, including:

Further, when the camera captures a dynamic target object, the target object is within a preset distance range of the monitoring image center of the camera.

Further, the camera is a 360-degree rotating high-definition night vision dome camera.

Further, drawing a target object pose picture through the continuous frame pictures, including:

Further, performing algorithm processing on the continuous frame pictures by adopting an object posture recognition algorithm, wherein the algorithm processing comprises the following steps:

Further, performing continuous dynamic recognition analysis and motion prediction on the target object posture picture to analyze the real behavior of the target object, including:

calculating a core feature vector of the key feature points;

Further, the convolutional neural network is a feature model obtained through deep learning of a plurality of human body posture pictures, and the feature model comprises human body bones, human face expressions, actions and posture feature models.

According to the behavior analysis device for the target object, when the image is shielded and blurred, the operations of target monitoring, trajectory drawing, alarming and the like can be continuously performed through the detection means of the millimeter wave radar, the usability of the system cannot be affected, and therefore the problem that monitoring work cannot be achieved by an intelligent video monitoring system under the condition that the image is blurred is solved.

Based on the same inventive concept, the present specification further provides an electronic device for analyzing target object behaviors, which includes at least one processor and a memory, where the memory stores a program and is configured to be executed by the at least one processor to:

drawing a target object posture picture through the continuous frame pictures;

For other functions of the processor, reference may also be made to the contents described in the above embodiments, which are not described in detail herein.

Based on the same inventive concept, embodiments of the present specification further provide a computer-readable storage medium including a program for use in conjunction with an electronic device, the program being executable by a processor to perform the steps of:

drawing a target object posture picture through the continuous frame pictures;

In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (e.g., improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD) (e.g., a Field Programmable Gate Array (FPGA)) is an integrated circuit whose Logic functions are determined by a user programming the Device. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually making an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abel (advanced Boolean Expression Language), ahdl (alternate Hardware Description Language), traffic, pl (core universal Programming Language), HDCal (jhdware Description Language), lang, Lola, HDL, laspam, hardward Description Language (vhr Description Language), vhal (Hardware Description Language), and vhigh-Language, which are currently used in most common. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the units may be implemented in one or more software and/or hardware when implementing the present application.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that 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 an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The application may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

The above description is only an example of the present application and is not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for behavior analysis of a target object, the method comprising:

drawing a target object posture picture through the continuous frame pictures;

2. The method of claim 1, wherein the method further comprises:

3. The method of claim 1, wherein the target area comprises a defense area, a pre-warning area, and an alarm area;

4. The method of claim 1, wherein the target trajectory information includes target coordinates, target speed, target angle, target distance.

5. The method of claim 1, wherein configuring a camera based on the target trajectory information comprises:

6. The method of claim 1, wherein the target object is within a preset distance range of a center of a surveillance image of the camera when the camera captures a dynamic target object.

7. The method of claim 5 or 6, wherein the camera is a 360 degree rotating high definition night vision dome camera.

8. The method of claim 1, wherein depicting a target object pose picture by the successive frame pictures comprises:

9. The method of claim 8, wherein the algorithmic processing of the successive frame pictures using an object pose recognition algorithm comprises:

10. The method of claim 1, wherein performing continuous dynamic recognition analysis and motion prediction on the target object pose picture to analyze the real behavior of the target object comprises:

calculating a core feature vector of the key feature points;

11. The method of claim 10, wherein the convolutional neural network is a feature model obtained by deep learning of a plurality of human pose pictures, the feature model comprising human bone, facial expression, motion and pose feature models.

12. The method of claim 10, wherein the key feature points comprise joints with degrees of freedom on the human body.

13. The method of claim 12, wherein the joint comprises a neck, a shoulder, an elbow, a wrist, a waist, a knee, an ankle.

14. An apparatus for behavior analysis of a target object, the apparatus comprising:

15. The apparatus of claim 14, wherein the method further comprises:

16. The apparatus of claim 14, wherein the target area comprises a defense area, a pre-warning area, and an alarm area;

17. The apparatus of claim 14, wherein the target trajectory information comprises target coordinates, target speed, target angle, target distance.

18. The apparatus of claim 14, wherein configuring a camera based on the target trajectory information comprises:

19. The apparatus of claim 14, wherein when the camera captures a dynamic target object, the target object is within a preset distance range of a center of a surveillance image of the camera.

20. The apparatus of claim 18 or 19, wherein the camera is a 360 degree rotating high definition night vision dome camera.

21. The apparatus of claim 14, wherein depicting a target object pose picture by the successive frame pictures comprises:

22. The apparatus of claim 21, wherein the algorithmic processing of the successive frame pictures using an object pose recognition algorithm comprises:

23. The apparatus of claim 14, wherein performing continuous dynamic recognition analysis and motion prediction on the target object pose picture to analyze the real behavior of the target object comprises:

calculating a core feature vector of the key feature points;

24. The apparatus of claim 23, wherein the convolutional neural network is a feature model obtained by deep learning of a plurality of human pose pictures, the feature model comprising human bone, facial expression, motion and pose feature models.

25. An electronic device for behavioral analysis of a target object, comprising at least one processor and a memory, the memory storing a program and configured for the at least one processor to perform the steps of:

drawing a target object posture picture through the continuous frame pictures;

26. A computer readable storage medium having stored thereon computer readable instructions executable by a processor to implement a method of behavioral analysis of a target object according to any one of claims 1 to 13.