阅读说明：本技术 一种基于视频ai的陌生人通行监控方法及装置 (Stranger passage monitoring method and device based on video AI ) 是由 容典 高俊 刘润成 于 2021-07-01 设计创作，主要内容包括：本发明提供了一种基于视频AI的陌生人通行监控方法及装置,包括如下步骤：S1：视频监控器通过图像采集模块对通行人群进行拍摄,获取多人通行图像视频数据；S2：计算机处理设备通过人脸识别模块对多人通行图像视频数据进行人脸识别,获取通行人群的人脸特征数据；S3：计算机处理设备通过视频AI模块对通行人群的人脸特征数据进行视频AI分析,判断通行人群中是否出现陌生人情况；S4：当检测到异常陌生人时,计算机处理设备通过陌生人预警模块向所述报警设备发送报警指令。该种基于视频AI的陌生人通行监控方法及装置,可以使监控摄像头智能地对陌生人进出的通行情况进行处理,保证小区、住宅楼等的用户安全,方便了治安管理。(The invention provides a stranger passing monitoring method and device based on video AI, comprising the following steps: s1: the video monitor shoots passing people through the image acquisition module to obtain multi-person passing image video data; s2: the computer processing equipment carries out face recognition on the multi-person passing image video data through a face recognition module to obtain face feature data of passing people; s3: the computer processing equipment carries out video AI analysis on the face characteristic data of the passing people through a video AI module to judge whether stranger conditions occur in the passing people; s4: when an abnormal stranger is detected, the computer processing equipment sends an alarm instruction to the alarm equipment through the stranger early warning module. The stranger passing monitoring method and device based on the video AI can enable the monitoring camera to intelligently process passing conditions of passing of strangers, ensure the safety of users in a community, a residential building and the like, and facilitate public security management.)

1. A stranger passing monitoring method based on video AI is characterized by comprising the following steps:

s1: the video monitor shoots passing people through the image acquisition module to obtain multi-person passing image video data;

s2: the computer processing equipment carries out face recognition on the multi-person passing image video data through a face recognition module to obtain face feature data of passing people;

s3: the computer processing equipment carries out video AI analysis on the face characteristic data of the passing people through a video AI module to judge whether stranger conditions occur in the passing people;

s4: when an abnormal stranger is detected, the computer processing equipment sends an alarm instruction to the alarm equipment through the stranger early warning module.

2. The stranger passing monitoring method based on the video AI as claimed in claim 1, wherein the video monitor shoots passing people through an image acquisition module to obtain multi-person passing image video data, and the method comprises the following steps:

according to the period before the passer-by passes through the video monitor, the video monitor acquires the image acquired after shooting the passer-by passing through the period so as to obtain the video data of the multi-person passing image;

the passers-by includes an acquaintance database of known real-identity persons entered into a database of the computer processing device and a stranger database of unknown stranger-identity persons not entered into the database.

3. The stranger passage monitoring method based on video AI of claim 1, wherein the computer processing device performs face recognition on the multi-person passage image video data through a face recognition module to obtain the face feature data of the passage people, and the method comprises the following steps:

transmitting the video data of the multi-person traffic image to a database of the computer processing equipment, and retrieving and corresponding the face data of each traffic person identified in the captured and collected images of the traffic persons acquired by the video monitor;

if the person is not searched in the database of the computer processing equipment, the person is marked as a stranger and is input into the stranger database of the computer processing equipment.

4. The stranger traffic monitoring method based on video AI of claim 1, wherein the computer processing device performs video AI analysis on the facial feature data of the traffic crowd through a video AI module to determine whether a stranger condition occurs in the traffic crowd, and the method comprises the following steps:

extracting a human face digital image in the multi-person passing image video data, preprocessing the human face digital image to be recognized, eliminating a background, and acquiring a face region in the human face digital image;

and recording geometrical characteristic values of five sense organs of the human body in the face area, including two eyes, two eyebrows, the distance between two ears, the height of a nose and the width of a mouth, and marking each geometrical characteristic value.

5. The method for monitoring the passage of strangers based on video AI of claim 4, wherein the geometric feature values further comprise the shapes of eyes, eyebrows, ears, nose and mouth.

6. The stranger passage monitoring method based on the video AI as claimed in claim 1, wherein when an abnormal stranger is detected, the computer processing device sends an alarm instruction to the alarm device through the stranger early warning module, comprising:

when the computer processing equipment does not retrieve the face feature data of the passing people in the acquaintance database and marks the passing people as strangers, alarm text information is generated and sent to the alarm equipment to give an alarm, and the recorded face feature data of the stranger are transmitted to the alarm equipment for being inquired by security personnel.

7. A stranger passing monitoring device based on video AI comprises a video monitor (1), a computer processing device (2) and an alarm device (3), and is characterized in that,

the method comprises the following steps: the image acquisition module is used for acquiring a digital image of the face of a person in the video information shot by the video monitor (1) and shooting time;

the face recognition module is used for recognizing and analyzing all faces in the digital face image of the person collected by the image collection module to obtain the face characteristics corresponding to each person;

the video AI module is used for storing the face features acquired by the face recognition module;

and the stranger early warning module is used for retrieving all recorded face characteristics stored by the face recognition module through the video AI module according to the face characteristics acquired by the face recognition module, and sending early warning through the warning device (3) if no corresponding face record is found.

8. The stranger passing monitoring device based on video AI according to claim 7, characterized in that the video monitor (1) comprises a shell (101) and a lens (102), the front inner surface of the shell (101) is embedded with a water mist removing component (4) which can remove the water mist on the surface of the lens (102) near the lens (102);

the water mist removing assembly (4) comprises an inner sleeve assembly (401).

9. The stranger passage monitoring device based on video AI of claim 8, wherein the inner sleeve assembly (401) comprises a plurality of channels (4011) and blocks (4012).

10. The stranger passage monitoring device based on the video AI of claim 9, wherein each of the blocks (4012) has a trapezoidal appearance in a longitudinal section, and the blocks (4012) are quicklime.

Technical Field

The invention relates to the field of computers, in particular to a stranger passing monitoring method and device based on video AI.

Background

The monitoring action of the traditional monitoring camera can be executed only by manual operation, and the camera around the moving target can not be automatically controlled to accurately monitor when the moving target is found.

Aiming at the problems, regarding the technical problem that the monitoring camera cannot automatically move, a monitoring device and a method for automatically focusing a plurality of monitoring cameras with the patent number of CN202010383904.2 are inquired through a large number of searches, a monitoring system is provided with m cameras, when a certain camera monitoring module finds that a moving target exists in a monitoring area, the distance di from the certain camera monitoring module to the monitoring camera and the size information of the moving target are measured, and then the communication module is utilized to realize the mutual sharing of the information of other cameras in the communication distance; each camera determines whether the camera is the front n (n is less than m) cameras closest to the moving target object according to the obtained distance information and an algorithm; executing a movement control instruction and a focusing instruction by the front n cameras closest to the moving target, and keeping the working states of other cameras unchanged; when the moving target moves out of the monitoring area, the camera returns to the initial state; the patent provides a monitoring device and a method for automatic focusing of a plurality of monitoring cameras, so that after a moving target is found by the cameras, a monitoring system can automatically distribute monitoring tasks to obtain picture information of the moving target more clearly.

However, the technical scheme provided by the patent can not process the passing condition that strangers pass in and out intelligently by the monitoring camera, so that the safety of users in a community, a residential building and the like is threatened, and the difficulty in security management is caused; and existing video monitors are susceptible to rain and fog in rainy and foggy weather.

Disclosure of Invention

The invention aims to solve the technical problems in the background art and provides a stranger passing monitoring method and device based on video AI.

In order to achieve the purpose, the invention provides the following technical scheme: a stranger passing monitoring method based on video AI comprises the following steps:

s1: the video monitor shoots passing people through the image acquisition module to obtain multi-person passing image video data;

s2: the computer processing equipment carries out face recognition on the multi-person passing image video data through a face recognition module to obtain face feature data of passing people;

s3: the computer processing equipment carries out video AI analysis on the face feature data of the passing people through a video AI module to judge whether stranger conditions occur in the passing people;

s4: when an abnormal stranger is detected, the computer processing equipment sends an alarm instruction to the alarm equipment through the stranger early warning module.

Further preferred embodiments: video monitor passes through the image acquisition module and shoots current crowd, acquires many people current image video data, includes:

according to the period before the passer-by passes through the video monitor, the video monitor acquires the image acquired after shooting the passer-by passing through the period so as to obtain the video data of the multi-person passing image;

the passers-by includes an acquaintance database of known real-identity persons entered into a database of the computer processing device and a stranger database of unknown stranger-identity persons not entered into the database.

Further preferred embodiments: computer processing equipment passes through face identification module and carries out face identification to many people image video data that passes, obtains the facial feature data of passing crowd, includes:

transmitting the video data of the multi-person traffic image to a database of the computer processing equipment, and retrieving and corresponding the face data of each traffic person identified in the captured and collected images of the traffic persons acquired by the video monitor;

if the person is not searched in the database of the computer processing equipment, the person is marked as a stranger and is input into the stranger database of the computer processing equipment.

Further preferred embodiments: the computer processing equipment carries out video AI analysis to the facial feature data of the passing crowd through the video AI module, and judges whether stranger conditions occur in the passing crowd or not, and the method comprises the following steps:

extracting a human face digital image in the multi-person passing image video data, preprocessing the human face digital image to be recognized, eliminating a background, and acquiring a face region in the human face digital image;

and recording geometrical characteristic values of five sense organs of the human body in the face area, including two eyes, two eyebrows, the distance between two ears, the height of a nose and the width of a mouth, and marking each geometrical characteristic value.

Further preferred embodiments: the geometric characteristic values also include the shapes of eyes, eyebrows, ears, nose and mouth.

Further preferred embodiments: when an abnormal stranger is detected, the computer processing equipment sends an alarm instruction to the alarm equipment through the stranger early warning module, and the alarm instruction comprises the following steps:

when the computer processing equipment does not retrieve the face feature data of the passing people in the acquaintance database and marks the passing people as strangers, alarm text information is generated and sent to the alarm equipment to give an alarm, and the recorded face feature data of the stranger are transmitted to the alarm equipment for being inquired by security personnel.

Further preferred embodiments: a stranger passing monitoring device based on video AI comprises a video monitor, a computer processing device and an alarm device, and is characterized in that,

the method comprises the following steps: the image acquisition module is used for acquiring a digital image of the face of a person in the video information shot by the video monitor and shooting time;

the face recognition module is used for recognizing and analyzing all faces in the digital face image of the person collected by the image collection module to obtain the face characteristics corresponding to each person;

the video AI module is used for storing the face features acquired by the face recognition module;

and the stranger early warning module is used for retrieving all recorded face characteristics stored by the face recognition module through the video AI module according to the face characteristics acquired by the face recognition module, and sending early warning through the warning equipment if corresponding face records are not found.

Further preferred embodiments: the video monitor comprises a shell and a lens, wherein a water mist removing component capable of removing water mist on the surface of the lens is embedded on the inner surface of the front side of the shell, which is close to the position of the lens;

the water mist removing component comprises an inner sleeve component.

Further preferred embodiments: and a rain shield is fixedly arranged at the top end of the shell.

Further preferred embodiments: the inner sleeve component comprises an inner groove, a plurality of channels and clamping blocks.

Further preferred embodiments: the improved shell is characterized in that an inner groove is formed in the front-back horizontal direction central position of the inner surface of the front side of the shell, the appearance of the inner groove is in a hollow cylindrical shape, a plurality of passages are arranged in an equal-adjacent row in the bottom to the upper position of the inner part of the inner groove in a penetrating mode, clamping blocks are embedded in the inner groove between every two adjacent passages, the appearance of each clamping block is in a trapezoid shape on one longitudinal section, and the components of each clamping block are quicklime.

Has the advantages that:

according to the stranger passing monitoring method and device based on the video AI, the passing condition of passing of strangers can be intelligently processed by the monitoring camera through the cooperative action of the image acquisition module, the face recognition module, the video AI module, the stranger early warning module and the like, the safety of users in a community, a residential building and the like is guaranteed, and the public security management is facilitated; through the setting of removing the water smoke subassembly, can prevent that video monitor from suffering from the invasion and attack of rain and fog in rain and fog weather.

Drawings

FIG. 1 is a schematic flow chart of a stranger passage monitoring method based on video AI in the invention;

FIG. 2 is a block diagram of the stranger passage monitoring device based on video AI in the invention;

FIG. 3 is a schematic overall structure diagram of the stranger passage monitoring device based on video AI according to the present invention;

FIG. 4 is a schematic diagram of the overall structure of a video monitor according to the present invention;

FIG. 5 is a schematic cross-sectional view of a video monitor according to the present invention;

FIG. 6 is an enlarged view of the structure at A in FIG. 5 according to the present invention;

FIG. 7 is a schematic view of the overall structure of the inner sleeve assembly of the present invention;

in FIGS. 1-7: 1-a video monitor; 101-a housing; 102-a lens; 103-rain shield; 2-a computer processing device; 3-an alarm device;

4-removing the water mist component; 401-inner sleeve component; 4011-channel; 4012-fixture block.

Detailed Description

The technical solution in the embodiment of the present invention will be clearly and completely described below with reference to fig. 1 to 7 in the embodiment of the present invention.

Example 1

Referring to fig. 1, in an embodiment of the present invention, a method for monitoring the passage of strangers based on a video AI includes the following steps:

s1: the video monitor 1 shoots passing people through an image acquisition module to obtain multi-person passing image video data;

s2: the computer processing equipment 2 carries out face recognition on the multi-person passing image video data through a face recognition module to obtain face feature data of passing people;

s3: the computer processing equipment 2 carries out video AI analysis on the face feature data of the passing people through a video AI module to judge whether stranger conditions occur in the passing people;

s4: when an abnormal stranger is detected, the computer processing device 2 sends an alarm instruction to the alarm device 3 through the stranger early warning module.

Example 2

Referring to fig. 1-2, the embodiment of the present invention differs from embodiment 1 in that: video monitor passes through the image acquisition module and shoots current crowd, acquires many people current image video data, includes:

according to the period before the passer-by passes through the video monitor, the video monitor acquires the image acquired after shooting the passer-by passing through the period so as to obtain the video data of the multi-person passing image;

the passers-by includes an acquaintance database of known real-identity persons entered into a database of the computer processing device and a stranger database of unknown stranger-identity persons not entered into the database.

In the embodiment of the present invention, the computer processing device performs face recognition on the multi-person passing image video data through the face recognition module to obtain the face feature data of the passing people, including:

transmitting the video data of the multi-person traffic image to a database of the computer processing equipment, and retrieving and corresponding the face data of each traffic person identified in the captured and collected images of the traffic persons acquired by the video monitor;

if the person is not searched in the database of the computer processing equipment, the person is marked as a stranger and is input into the stranger database of the computer processing equipment.

In the embodiment of the present invention, the computer processing device performs video AI analysis on the face feature data of the passing people through the video AI module to determine whether a stranger condition occurs in the passing people, including:

extracting a human face digital image in the multi-person passing image video data, preprocessing the human face digital image to be recognized, eliminating a background, and acquiring a face region in the human face digital image;

and recording geometrical characteristic values of five sense organs of the human body in the face area, including two eyes, two eyebrows, the distance between two ears, the height of a nose and the width of a mouth, and marking each geometrical characteristic value.

In the embodiment of the invention, the geometric characteristic values further comprise the shapes of eyes, eyebrows, ears, noses and mouths.

In the embodiment of the present invention, when an abnormal stranger is detected, the sending, by the computer processing device, an alarm instruction to the alarm device through the stranger early warning module includes:

when the computer processing equipment does not retrieve the face feature data of the passing people in the acquaintance database and marks the passing people as strangers, alarm text information is generated and sent to the alarm equipment to give an alarm, and the recorded face feature data of the stranger are transmitted to the alarm equipment for being inquired by security personnel.

According to the stranger passing monitoring method and device based on the video AI, the passing condition of passing of strangers can be intelligently processed by the monitoring camera through the cooperative action of the image acquisition module, the face recognition module, the video AI module, the stranger early warning module and the like, the safety of users in a residential district, a residential building and the like is guaranteed, and public security management is facilitated.

Example 3

Referring to fig. 3-7, the embodiment of the present invention differs from embodiment 1 in that: a stranger passing monitoring device based on video AI comprises a video monitor 1, a computer processing device 2 and an alarm device 3, and is characterized in that,

the method comprises the following steps: the image acquisition module is used for acquiring a digital image of the face of a person in the video information shot by the video monitor 1 and shooting time;

the face recognition module is used for recognizing and analyzing all faces in the digital face image of the person collected by the image collection module to obtain the face characteristics corresponding to each person;

the video AI module is used for storing the face features acquired by the face recognition module;

and the stranger early warning module is used for retrieving all recorded face characteristics stored by the face recognition module through the video AI module according to the face characteristics acquired by the face recognition module, and sending early warning through the warning device 3 if no corresponding face record is found.

In the embodiment of the invention, the video monitor 1 comprises a shell 101 and a lens 102, wherein a water mist removing component 4 capable of removing water mist on the surface of the lens 102 is embedded in the position, close to the lens 102, of the inner surface of the front side of the shell 101;

the water mist removing assembly 4 comprises an inner sleeve assembly 401.

In the embodiment of the invention, a rain shield 103 is fixedly arranged at the top end position of the shell 101;

here, the rain shield 103 can protect the video monitor 1 from rain and fog in outdoor rain and fog weather.

In the embodiment of the present invention, the inner sleeve assembly 401 includes an inner groove, a plurality of channels 4011 and a plurality of blocks 4012.

In the embodiment of the invention, an inner groove is formed in the front-back horizontal direction center position of the inner surface of the front side of a shell 101, the appearance of the inner groove is in a hollow cylindrical shape, a plurality of passages 4011 are arranged in an equal-adjacent row at the bottom to the upper end of the inner part of the inner groove, fixture blocks 4012 are embedded in the inner groove between every two adjacent passages 4011, the appearance of each fixture block 4012 is in a trapezoid shape on a longitudinal section, and the components of each fixture block 4012 are quicklime;

the channel 4011 can facilitate the water mist on the surface of the shell 101 to flow down through the channel 4011 to cause a subsequent trigger reaction; the composition here is lime and the outward appearance is trapezoidal fixture block 4012 on a longitudinal section, can make things convenient for the water smoke to flow down the back through passageway 4011, can flow to fixture block 4012 surface, produces the high temperature heat, and the high temperature heat can reverse return to lens 102 surface, dries to get rid of water smoke.