Infant behavior monitoring and alarming system and methodTechnical Field
The invention belongs to the field of computer vision, and particularly relates to an infant behavior monitoring and alarming system and method.
Background
Driven by wide application requirements, research on capturing and analyzing human motion through videos is more and more, but currently, a lot of problems still exist in directly and automatically acquiring human motion data from videos. The main difficulties are the high complexity from the human body itself and its movements and the inherent loss of information during camera imaging; in addition, the human body in the video may be frequently covered or self-shielded, and the resolution of the camera for acquiring the video is limited in time and space, and these objectively existing factors also make the tracking and acquiring of the human body motion data from the video sequence still a difficult technical challenge so far, and these difficulties are also the technical key of the human body motion analysis research.
Disclosure of Invention
The invention aims to solve the problems in the prior art and provides a system and a method for monitoring and alarming children behavior, which track the behavior activities of children by utilizing a human body joint motion characteristic tracking technology, identify and alarm the behavior activities such as window climbing, touch panel, gas stove playing and the like, ensure that the children can play safely at home independently, keep away from danger, worry parents and provide guarantee for the home.
The invention is realized by the following technical scheme:
a plurality of network cameras are arranged in a space, and whether a human body is close to a dangerous area or not is judged by processing and analyzing video images acquired by the network cameras.
The operation of arranging a plurality of network cameras in a space includes:
dividing a plurality of dangerous areas in the space, and setting the range of each dangerous area;
a plurality of network cameras are arranged in the space, so that the monitoring range of the network cameras can cover all dangerous areas.
The operation of judging whether the human body is close to the dangerous area by processing and analyzing the video images acquired by the network cameras comprises the following steps:
(1) image acquisition: collecting video images of all the network cameras in real time;
(2) judging whether the human body is collected in each video image or not by using the freedom degree of the human body joint motion model, if so, turning to the step (3), and if not, returning to the step (1);
(3) calculating the distance between the human body and the dangerous area, judging whether the human body is close to the dangerous area, if so, turning to the step (4), and if not, returning to the step (1);
(4) and sending an alarm signal.
The operation of judging whether the human body is collected in each video image by using the degree of freedom of the human body joint motion model in the step (2) comprises the following steps:
and judging whether motion characteristics of 7 rotational degrees of freedom of the human body joint motion model are extracted from the video images acquired by each network camera, if so, judging that the human body is acquired from the video images, and if not, judging that the human body is not acquired from the video images.
The operation of calculating the distance between the human body and the dangerous area in the step (3) comprises:
and carrying out image processing on the collected video image of the human body, and calculating the pixel distance between the human body and the dangerous area in the processed image, wherein the pixel distance is the distance between the human body and the dangerous area.
The operation of judging whether the human body approaches to the dangerous area in the step (3) comprises:
and judging whether the distance between the human body and the dangerous area in the video images collected by at least two network cameras is smaller than a set value, if so, judging that the human body is close to the dangerous area, and if not, judging that the human body is not close to the dangerous area.
The alarm signal in the step (4) comprises: voice alarm, light alarm, short message alarm, mail alarm, and dialing alarm call.
The step (4) further comprises:
video images in the time periods before and after the human body approaches the dangerous area are additionally backed up.
An infant behavior monitoring alarm system for implementing the method comprises: the system comprises a control center host and a plurality of network cameras;
each network camera is respectively connected with a control center host;
the control center host is provided with an image acquisition module, a human body detection module, a distance measuring and calculating module and an alarm prompting module;
the image acquisition module acquires video images transmitted by all the network cameras in real time and transmits the video images to the human body detection module;
the human body detection module judges whether a human body is collected or not by using the freedom degree of the human body joint motion model, and if so, sends a measurement and calculation starting signal to the distance measurement and calculation module;
after receiving the measurement starting signal, the distance measurement module calculates the distance between the human body and the dangerous area in the video image, judges whether the human body is close to the dangerous area according to the distance between the human body and the dangerous area, and sends an alarm signal to the alarm prompting module if the human body is close to the dangerous area;
and the alarm prompt module sends an alarm signal.
Preferably, the lens of the network camera adopts a camera lens with an annular infrared illumination function.
Compared with the prior art, the invention has the beneficial effects that:
the invention realizes the monitoring of the infant by utilizing the human body joint movement characteristic tracking technology, parents can open a computer or a mobile phone at any time to check the real-time images at home after the family is in operation or when the parents go out, communicate with the child face to face, know the family condition, remotely take care of the family, and can remotely assist in handling the emergency, so that the parents can feel more relieved to work, and the invention can completely realize the following functions:
1. remote monitoring
Essentially all surveillance systems support some form of internet remote access, the most prominent difference being the ease with which a public internet access video surveillance system can be accessed.
2. Audio monitoring
Often for important members of the family, such as young parents, young children, or pets, the audio monitoring function may enable the family user to see and hear their activities. The only practical way to achieve audio monitoring and at a relatively reasonable cost is to use a webcam at home.
3. Low light environment monitoring
Many inexpensive cameras do have a rather poor imaging effect in low light environments. The least costly and most efficient approach is to use a camera lens with an annular infrared illumination function. Although such devices are unbalanced in lighting and have limited monitoring distances, which are a great disadvantage for professional security monitoring applications, they are generally inexpensive and good for monitoring individual room effects.
4. Multiple camera positioning
According to the invention, whether danger exists or not is judged and an alarm is given out by directly calculating the linear distance between the dangerous area and the tracked person, the blind spot problem of a single camera is considered, the accuracy is improved by positioning multiple cameras, and an alarm is given out when multiple cameras judge that the infant is close to the same dangerous area.
Drawings
FIG. 1 is a block diagram of the steps of the method of the present invention;
FIG. 2 is a block diagram of the system of the present invention.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings:
the invention provides a solution for improving the safety guarantee of children by utilizing a computer graphic image analysis processing technology and a pattern recognition technology, automatically recognizes the dangerous behaviors of the children by monitoring through a camera and gives an alarm to contact parents, and a remote language system can lead the parents to converse with the children in real time and carry out safety education in time so as to prevent the reoccurrence of the dangerous behaviors.
The method of the invention is shown in figure 1 and comprises:
a plurality of network cameras are arranged in the space, and the monitoring range of the network cameras can cover all dangerous areas. A plurality of dangerous areas are divided in space, and the range of each dangerous area is set. The dangerous area refers to any area with potential safety hazard, such as the periphery of various dangerous goods, and then the following steps are carried out:
(1) image acquisition: collecting video images of all the network cameras in real time;
(2) judging whether motion characteristics of 7 rotational degrees of freedom of the human body joint motion model are extracted from the video images acquired by each network camera, if so, judging that the human body is acquired from the video images, turning to the step (3), otherwise, judging that the human body is not acquired from the video images, and returning to the step (1);
(3) calculating the distance between the human body and the dangerous area: the method comprises the steps of carrying out image processing on a collected video image of a human body, calculating a pixel distance between the human body and a dangerous area (the pixel distance between the center or the edge of the human body and the center of the dangerous area, the pixel distance between the center or the edge of the human body and the edge of the dangerous area, and specifically determining the pixel distance between the two points according to actual needs), wherein the pixel distance is the distance between the human body and the dangerous area. Then judging whether the pixel distance between the human body and the dangerous area in the video images of at least two network cameras is smaller than a set value, if so, judging that the human body is close to the dangerous area, then turning to the step (4), and if not, returning to the step (1);
(4) sending an alarm signal, the alarm signal comprising: voice alarm, light alarm, short message alarm, mail alarm, dialing alarm call, etc.
The system for realizing the method comprises the following steps:
1. the network camera: and shooting and monitoring the whole indoor environment in real time by deploying a plurality of network cameras, and assigning a dangerous area.
2. The control center host computer: the human body joint motion model has 7 rotational degrees of freedom, and the human body tracking is realized by extracting the motion characteristics of the 7 rotational degrees of freedom (the specific method can refer to the design and realization of a human body joint motion characteristic tracking platform in the reference of the patent document of Yuxiawei, the modern electronic technology of 2016 and 39 (22)).
As shown in fig. 2, an image acquisition module, a human body detection module, a distance measurement module and an alarm prompt module are arranged on the control center host; the image acquisition module acquires video images transmitted by each network camera and transmits the video images to the human body detection module; the human body detection module judges whether a human body is collected or not by using the freedom degree of the human body joint motion model, and if so, sends a measurement and calculation starting signal to the distance measurement and calculation module; after receiving the measurement starting signal, the distance measurement module calculates the distance between the human body and the dangerous area in the video image, judges whether the human body is close to the dangerous area according to the distance between the human body and the dangerous area, and sends an alarm signal to the alarm prompting module if the human body is close to the dangerous area; the alarm prompting module sends an alarm signal to a guardian and a related mechanism. The control center host is integrated with a voice system, and can be configured to send out voice alarm or not, select alarm sound and the like.
The process that the distance measuring and calculating module judges whether the human body approaches the dangerous area or not according to the distance between the human body and the dangerous area is as follows: when the distances between the human bodies in the video images acquired by the plurality of network cameras and the same dangerous area are smaller than a set value, the fact that the children are approaching the dangerous area can be judged. The dangerous area is designated and marked when the network cameras are deployed, each network camera shoots in real time, video images are transmitted to the control center host in real time, the control center host carries out image analysis, when the distance between a human body and the dangerous area is smaller than a set value, danger is possible, and when the distances between the human body in the video images acquired by the plurality of network cameras and the same dangerous area are smaller than the set value, the danger can be confirmed. After the danger is confirmed, the alarm prompt module on the control center host informs parents through a mobile phone, and the parents can connect a voice system in a house to communicate with children.
One embodiment of the invention is as follows:
aiming at the particularity of home video monitoring, the safety of the whole home range is guaranteed, and the problems of the two aspects are mainly solved by the embodiment, namely monitoring of the internal conditions of the home including video, audio and the like, and remote real-time monitoring, which is mainly realized by providing a broadband network by an operator at present. It is recommended to install two home network cameras per home.
1. Aspects of video surveillance
During the time from the target object entering the sight range of the network camera to the target object leaving, all the actions of the object are clearly transmitted to the control center host, the host can analyze the image, judge the alarm and send the alarm, and the host can be informed through the network, so that reliable evidence is provided for the scheduling of the duty center. The control center host is connected with a storage device, and the storage device can adopt an embedded hard disk video recorder and is used for storing video images transmitted by all the network cameras.
Secondly, a network camera is installed near the dangerous area, for example, a household network camera is installed in a living room and a balcony of the living room, the installation position is in the corner of the living room, and the whole picture is required to be adjusted in the monitoring range so as to monitor the access door of the balcony and most of the area of the living room. The network camera is arranged at the doorway and other positions with hidden dangers, such as windows of a kitchen or a toilet, the installation position is opposite to the dangerous area, and the dangerous area can be monitored completely.
And thirdly, the network camera is a commercially available household network camera, a tripod head is arranged in the network camera, and the network camera can rotate in the horizontal direction and the vertical direction, so that the monitoring range is increased to a great extent.
2. Fire prevention and theft prevention aspect
Firstly, an alarm probe is arranged at a key position, and comprises a door magnet, an infrared detector, a vibration detector, a smoke detector, a gas monitoring detector and the like, and the detectors are connected to an alarm interface of a network camera.
Secondly, in alarm linkage, the invention can realize the following functions: alarming by siren sound and lamplight inside a family; the alarm event is sent to a mail in time to inform the user (a mailbox can be bound on a control center host to automatically send the mail, and matched app and app prompt can be installed on a mobile phone of the user or remote voice conversation can be realized through the app, which can be realized by adopting the prior art); and functions of automatic alarm snapshot, automatic alarm video recording and the like can be realized (when a danger is judged, the control center host can additionally back up videos before and after a dangerous time period).
And thirdly, in the alarm function, the control center host has the functions of transmitting the alarm event to the outside in time, such as timely informing the community management, timely dialing the alarm call, timely sending the short message to the preset mobile phone and the like.
3. Video storage aspect
The video and the image can be stored on the SD card built in the camera, and also can be stored on other computer hard disks.
4. Aspect of realizing remote video monitoring
Each network camera is accessed to a router in a home by using a wireless or wired network, and is accessed to a broadband network of an operator through the router to transmit a video to a public network, so that remote monitoring is realized. Usually, a domain name is requested and bound to the router, so that the video image of the home can be conveniently accessed at a remote place. The network camera transmits the video to the control center host, the control center host transmits the video to the public network, and remote monitoring is also used for synchronously checking image video information acquired by the control center host.
The above-described embodiment is only one embodiment of the present invention, and it will be apparent to those skilled in the art that various modifications and variations can be easily made based on the application and principle of the present invention disclosed in the present application, and the present invention is not limited to the method described in the above-described embodiment of the present invention, so that the above-described embodiment is only preferred, and not restrictive.