RU2268497C2 - System and method for automated video surveillance and recognition of objects and situations - Google Patents

Abstract

FIELD: engineering of alarm signaling equipment.

SUBSTANCE: video surveillance system has video camera, computing device with memory, target recognition device, moving video camera, database with previously recorded descriptions of alarming situations, realized on computing device with memory, and surveillance terminal. Moving video camera is made with possible tracking of target dependently on controlling signals received from target recognition device. Database is made with possible access thereto from moving video camera and from target recognition device and recording of new information, received from moving video camera. Surveillance terminal is engineered for informing operator and receiving controlling commands from operator. Target recognition device is engineered for analyzing signal, received from video camera with possible recognition of alarming situation and detecting target in accordance to given algorithm with use of alarming situations description, stored in database. Target recognition device outputs appropriate alarming signal to moving video camera for tracking target and alarm signal - to surveillance terminal. Moving video camera is made with possible production of target image, useable for further analysis, and its transfer to database. Video surveillance method includes stages for receiving video signal by controlled zone video camera, its transmission and analysis in computing device with memory. Signal received by video camera is analyzed by computing device with use of information including alarming situations descriptions, prerecorded in form of database in memory of computing device. On basis of analysis results, data about target are produced, controlling signal is generated, and transferred to moving camera, capable of tracking target in accordance with controlling signal. By means of moving video camera, image of target is received, useable for further analysis, which is then recorded in database.

EFFECT: group of inventions is characterized by expanded functional capabilities.

2 cl, 2 dwg

Description

Technical field

The present invention relates to the field of video surveillance and object recognition and more specifically relates to a system and method for automated video surveillance and recognition of objects and situations.

State of the art

The prior art automated (hardware-software) systems and methods for video surveillance, recognition, identification and monitoring of objects intended for use in various fields. Most often, such systems are used to solve research problems in the field of technical safety systems, as well as for environmental monitoring and traffic control.

A known system according to the patent of the Russian Federation 2137203, designed to control the flow of vehicles, containing digital cameras installed at various points, computers and / or servers connected via a local area network. In the memory of computers there are databases with the possibility of their analysis and comparison with images received from video cameras. The system allows with a certain error to identify individual search objects in controlled areas and register them. There is a system interface with the operator through the monitor at the checkpoint. The limitations of this system are the limited control zones (cameras are fixed and have a limited viewing angle), and the cameras are static. In addition, this system does not provide for the possibility of recognizing the situation in the control zone.

A known video surveillance system according to US patent 5982420, containing video cameras with a rotary device and a variable focal length (buzzer) connected to a computer. This system provides for the automatic pointing of the camcorder to an object falling into the controlled area. The system registers such an object and dynamically monitors it. The system does not provide for the possibility of recognizing the situation and corresponding response to the situation, and also does not provide tracking of several objects.

A known monitoring and aiming system according to the published patent application of the Russian Federation No. 200212339, containing sequentially located and optically paired scanning device forming an optical device, a rotary optical device and a visual observation device (video camera). This system allows tracking the target in an automated mode. The system does not allow the analysis of the situation in the observation zone in accordance with predefined alarm situations and react accordingly.

SUMMARY OF THE INVENTION

The present invention is based on the technical task to create a system and method for automated video surveillance and recognition of situations that, in addition to recognizing, tracking and registering objects, would provide the following additional functions: recognize and classify objects according to certain signs in various lighting conditions, operational search for an object by specified parameters , recognition of the situation by the actions of objects, with the ability to track such objects, save in memory Istemi image suitable for subsequent analysis and issuing an alarm.

The specified technical problem is solved due to the fact that the proposed system (device) of automated video surveillance and recognition of objects and situations, containing, containing at least one video camera and at least one computing device with memory connected via a local area network, which according to the present invention contains at least one target recognition device connected to the video camera, at least one implemented on a computing device with memory, b a data base with descriptions of alarm situations pre-recorded in it, made with the possibility of access to it from the video camera and from the target recognition device and recording of new information received from the video camera into it, at least one movable video camera made with the ability to aim at the target in depending on the control signal received from the target recognition device, and at least one monitoring terminal adapted to inform the operator and receive control commands from him, the device target recognition is adapted to analyze a signal from a video camera with the ability to recognize an alarm situation and determine the target in accordance with a predetermined algorithm using the description of alarm situations stored in the database and issuing the corresponding control signal to a mobile video camera to aim at the target and an alarm signal to the terminal observation, and a movable video camera is configured to obtain a target image suitable for further analysis, and transmit it for storage in a database.

Preferably, in the system according to the present invention, the target recognition device is implemented in software in a computing device with memory. The movable video camera is configured to rotate in the vertical and horizontal plane, and / or change the focal length, and / or change the aperture and / or focus setting to provide guidance to the target. The system according to the invention advantageously further comprises a target detection sensor coupled to the target recognition device, the target recognition device adapted to analyze a signal from the target detection sensor. As a target detection sensor, a device selected from the group consisting of: infrared cameras, thermal imaging cameras, radar devices, a displacement sensor is used.

Preferably, the target recognition device is configured to calculate a velocity vector, the current coordinate of finding the target, generate and transmit to the mobile video camera the corresponding control signal to ensure that the mobile video camera is aimed at the target. The target recognition device is configured to transmit a control signal to the mobile video camera via a serial communication channel.

The predominantly specified algorithm provides the analysis of the signal from the video camera in accordance with the parameters selected from the group including brightness, size, duration of stay in the controlled area, speed and direction of movement. A personal computer was used as a computing device with memory. According to the invention, in the memory of the computing device, by means of software, a map model of the space controlled by the system can be formed, configured to apply a target model to it reflecting the position of the target in the space controlled by the system, in accordance with the data obtained by processing in the signal target recognition device obtained from the video camera, and this card model is adapted for analysis by the operator. The model map can be made three-dimensional.

The technical problem is also solved due to the fact that the proposed method of video surveillance, including the steps of receiving the video signal of the controlled area of at least one video camera, transmitting it and analyzing it in at least one computing device with memory, in which, according to the invention, a video camera the signal is analyzed using a computing device using information with descriptions of alarm situations, pre-recorded as a database in the memory of the computing device, according to the results of Isa obtained data on the purpose, a control signal is formed, it is transmitted to the mobile camera adapted for targeting in accordance with the control signal via the mobile camcorder lecture target image suitable for further analysis, and transmits it for storage in the database.

Preferably, in the method according to the invention, according to the results of the analysis, an alarm is additionally generated and transmitted to the operator. The analysis is carried out with the ability to calculate the velocity vector, the current coordinate of finding the target, and additionally generate and transmit to the mobile video camera the corresponding control signal to ensure that the moving video camera is aimed at the target. The control signal is transmitted to the mobile video camera via a serial communication channel. I analyze the signal from the video camera taking into account the target parameters selected from the group including brightness, size, duration of stay in the controlled area, speed and direction of movement. As a computing device with memory, a personal computer is mainly used. Preferably, in the memory of the computing device, by means of software, a map model of the space controlled by the system is formed with the possibility of applying a target model to it reflecting the position of the target in the space controlled by the system, in accordance with the data obtained by processing the signal received from the video camera in the target recognition device, This model card is adapted for analysis by the operator. The model card can be formed three-dimensional.

DETAILED DESCRIPTION OF THE INVENTION

The present invention will be clear to the specialist from the following more detailed description of the present invention, as well as a specific non-limiting example of its implementation with reference to the drawings, in which:

- figure 1 depicts a General functional diagram of a video surveillance system and recognition of objects and situations according to the present invention;

- figure 2 - schematically depicts a video surveillance system according to a specific non-limiting example implementation of the present invention.

The same type of device is indicated in the drawings by the same reference numbers.

Figure 1 presents the General functional diagram of a system of automated video surveillance and recognition of objects and situations (hereinafter the system). The system consists of interconnected devices electrically connected according to the rules known in the art. Separate devices that perform similar functions in the system are combined into logical blocks for ease of analysis in the circuit. This division into logical blocks is conditional and made for the convenience of system analysis and does not limit the essence of the invention.

The background analysis and target detection unit is indicated in FIG. 1 by 1. The unit 1 is a set of devices for detecting targets using various physical principles for this. Such devices can be used: color video cameras 11 (for example, WAT-220, available on the Russian market by Watec), black-and-white video cameras of camera 12 (for example, WAT-902H, from Watec), infrared cameras 13 (for example, TVC200ML, Central Research Institute "Elektron"), thermal imaging cameras 14 (for example, TV-04 Kst, f. NTPP NIIITsPP), radar devices 15 (for example, radar "ORWELL", f. GUP SPC "ELVIS"), motion sensors 16 and determining the coordinates of finding the target (for example, 525D, presented on the Russian market by Paradox). Block 1 may include a different number and combination of the above devices, depending on the specific needs of users of the system, as well as on operating conditions. Devices can be installed with duplication of each other's work and mutual overlapping of control zones to ensure increased reliability of the system. The devices of Unit 1 are electrically connected through an interface (for example, a composite video signal or Ethernet) with unit 2 for recognizing targets, determining coordinates, target parameters, and analyzing situations. The devices of Block 1 are adapted to transmit the received zone images or another target recognition signal to Block 2. Block 2 contains a device 21 for recognizing the target, determining coordinates and parameters of targets, and analyzing situations, which can be implemented as a specialized computing device (for example, an industrial computer based on PC-104 standard modules using the Multicor processor) or based on a universal personal computer (for example, with the Intel Pentium processor) equipped with programs th software allowing to receive and drive the digitized information from the background analysis devices and detection purposes; remember and track the state of the background, including its changes (time of day, state of the atmosphere - clear, cloudy, precipitation, season, and state of cover); calculate the velocity vector, the current coordinate of finding the target; save recognized parameters of the target to the database. Block 2 recognition of goals, determination of coordinates, parameters of goals and analysis of situations is connected electrically via an interface (for example, RS-485) with Block 3 of movable controlled cameras. Block 2 is adapted through the use of software, for example, typical in this area, for issuing control commands and transmitting the coordinates and velocity vector of the target to Block 3. Block 2 is also configured to control movable cameras 31 of Block 3 via a serial communication channel, due to the use of , for example, typical software. Block 2 is also adapted to analyze the signals received from the devices of Block 1 in accordance with a predetermined algorithm using the information stored in the Database of Block 4 with which Block 2 is connected. Block 2 is adapted for issuing, depending on the recognized situation, an alarm signal to terminals 51 for monitoring the analysis of results and control of the system. Block 2 is configured to issue an alarm to the terminals 51 and a control signal of the video cameras 31. Block 3 may contain a different number of movable controlled video cameras 31 of various types (for example, Panasonic WV-CS854 f. Panasonic), adapted to perform the following functions in accordance with received from Block 2 by the management team: rotation in the vertical and horizontal plane; change in focal length; focus setting; Changing the aperture or shutter speed. Block 3 mobile video cameras are intended for obtaining a detailed enlarged image of targets used for additional recognition of target parameters or detailed tracking of a target at the request of a system operator. Block 3 is connected with Block 4, which is a computer with a storage device (for example, a personal computer). In this storage device, a database has been created with the ability to store a video archive. The database of Block 4 is made with the possibility of automatically replenishing and storing information about all recognized goals, recognized situations, parameters of the goal, time of occurrence and loss of goals, photographs of goals, fragments of the video archive that reflect all the actions of the target when it is in the control zone of the video system. Enlarged images received by the mobile cameras of Unit 3 in the form of digitized photographs can be stored in the database of Unit 4. These photos can be used in the future to study the situation by a specialist and / or operator. Block 4 is electrically connected with Block 2 recognition of goals, determining coordinates, parameters of goals and analysis of situations. Block 4 and Block 2 are associated with the possibility of transmitting data to each other. Block 4 is also associated with Block 5 of the monitoring, analysis, and control terminals. Terminals 51 for monitoring, analyzing results and managing the system are one or more personal computers connected by communication networks. Terminals 51 for monitoring, analyzing the results, and controlling the system are configured to monitor what is happening in the control zone of the system (observation zone for background analysis and target detection devices). This is achieved through the use of, for example, a typical monitor (color or black and white) and typical software. The terminals 51 of Block 5 are configured to observe the situation on the screen and access the operator to control the system through a keyboard or other specialized person-computer interface (for example, a joystick).

Depending on the tasks facing the system of the present invention, various specific system configurations can be selected. In particular, logical units 2, 4, and 5 of the system can be implemented on various hardware. Blocks 2, 4 and 5 can be implemented on one specialized computing device (for example, an industrial computer based on modules of the PC-104 standard using the Multicor processor) or on one personal computer (for example, based on the Intel Pentium processor) as specially programmed memory areas. Blocks 2, 4 and 5 can also be made on the basis of a distributed network system, which is a combination of specialized computing devices and personal computers (with or without monitors). In general, the present invention encompasses various combinations and variants of hardware and software implementations of target recognition devices, determining coordinates and target parameters, analyzing situations, a database and video archive, observation terminals, analyzing results and managing the system. A specific hardware and software implementation is selected based on their system requirements in accordance with the rules known to a person skilled in the art.

The system and method of automated video surveillance and recognition of objects and situations according to the present invention will be clear to a person skilled in the art from the following description of the operation of a specific non-limiting example of the implementation of a system of automated video surveillance and recognition of objects and situations, which is carried out with reference to figure 2 of the drawings. This example system according to the present invention is illustrative and does not limit the set of essential features of the present invention.

The system of this particular embodiment of the invention includes two stationary video cameras 61 of the WAT-220 type and one infrared (IR) camera 62 of the TVC200ML type as background analysis and target detection devices, a universal personal computer 64 with an Intel Pentium processor, and a random access memory 256 kB bytes and a 20 MB hard disk as a device for target recognition, determining coordinates and parameters of targets, and analyzing situations. This personal computer has a database and a terminal for monitoring results analysis and system management. The system also includes a 63 WV-CS854 movable guided camcorder.

The system operates as follows. Video cameras 61 and an infrared camera 62 form images of two parts of the overall video scene of the system. Moreover, the images of parts of the video scene can touch one side (or be independent), forming the aggregate video scene on which the system operates. The IR camera 62 can form a separate video scene or form an aggregate image together with video cameras 61. This use of video cameras allows you to create an image in daylight from the image of the cameras 61, and at night from the image of the camera 62. Images are transmitted via the interface to the computer 64. Computer 64 in accordance with software based on the Windows operating system, it converts the composite analog video signal received from cameras into digital form and processes in the image, namely, it performs the following actions: it analyzes the background, detects goals, recognizes goals and situations, determines the parameters of the goals, their coordinates, remembers the motion paths, transfers the recognition results to the database, attaching events to time, records the video signal in compressed or raw form in the computer's memory in the form of files whose address is located in the database, thus, the video signal is associated with events, information about which is located in the database. The coordinates of the targets are transmitted to the mobile video camera 63 in the form of a set of target designations taking into account the prediction of the location of targets at the time of pointing the camera at the target. The computer 64 also digitizes the video image of the mobile video camera 63, the image can be broadcast on a computer screen, one of the frames is entered into the database as a detailed photograph of the target. The image is transmitted for processing in order to recognize the target and its parameters from its detailed image using static recognition algorithms known to a person skilled in the art.

Images of video cameras can be projected onto a plan or object map entered and stored by the system in the form of map areas covered with images of video cameras. Recognized objects in the form of mnemonic signs corresponding to a particular type of target are projected in real time onto a map or plan of the object in accordance with the coordinate of their location. The operator, if necessary, can mask any arbitrary part (or an arbitrary number of parts) of the image or part of the map of the object, at the same time, image areas freed from the mask can be formed inside the mask. In this case, events occurring in the mask area are ignored by the system. One or an arbitrary number of areas subject to increased protection, so-called, can also be formed on the image. alarm area. These areas may be of arbitrary shape, possibly having within themselves areas of arbitrary shape that are not an alarm zone. Each alarm zone can be assigned a name, time and date of operation, type and parameter of an object or situation. If an object occurs in the alarm zone with the specified parameters at a specified time or when a programmed situation occurs, the system will issue the appropriate message to the system operator’s terminal. The operator, having activated the alarm, automatically receives an image of the event that caused the alarm on the screen, regardless of whether it is happening now or has happened earlier for some time. An alarm occurs as an event in the database. The system operator can, if necessary, instruct the system to follow his chosen goal or several goals specified arbitrarily in the image, or by selecting several objects with certain parameters (for example, having the highest speed of movement, or all objects of the same type, or color, or the like). In this case, the system will automatically alternately display on the screen images of targets transmitted by the mobile video camera 63 in a detailed form.

The operator can train the system to recognize new types of targets. To do this, he repeatedly indicates the image of a specific target in the image and enters its name. In this case, the system is trained to recognize a specific new target.

Industrial applicability

The system and method according to the present invention are intended primarily for use in security and monitoring systems. The system can be implemented both on the basis of specialized devices (high-resolution video cameras, specialized radar devices, displacement sensors and computer systems), and on the basis of standard television cameras and general-purpose computers. Depending on the configuration of the system, as well as on ensuring duplication of the operation of individual devices, it is possible to achieve the required reliability and level of control over the observed area. Using specialized equipment and ensuring duplication increases the quality and reliability of work and leads to an increase in the cost of the system.

The system and method for automated video surveillance and recognition of objects and situations according to the present invention allows efficient automated monitoring of the protected area under various weather conditions and time of day, with a high degree of reliability to recognize programmed alarm situations, issue alarms and record the alarm situation in the form of a video suitable for further analysis by a specialist.

Claims (19)

1. A video surveillance system comprising at least one video camera and at least one computing device with a memory, characterized in that it contains at least one target recognition device connected to the video camera, at least one movable a video camera configured to aim at a target, depending on the control signal received from the target recognition device, at least one database implemented on a computing device with memory with a description alarms made with the possibility of access to it from a mobile video camera and from a target recognition device and recording new information received from a mobile video camera and at least one monitoring terminal adapted to inform the operator and receive control commands from him moreover, the target recognition device is adapted to analyze a signal from a video camera with the ability to recognize an alarm situation and determine the target in accordance with a predetermined algorithm using the description of the alarm situations stored in the database and the issuance of the corresponding control signal to the mobile video camera for aiming at the target and the alarm signal to the monitoring terminal, and the mobile video camera is configured to receive a target image suitable for further analysis and transmit it for storage in a database. 2. The CCTV system according to claim 1, characterized in that the target recognition device is implemented programmatically in a computing device with memory. 3. The CCTV system according to claim 1, characterized in that the movable video camera is configured to rotate in a vertical and horizontal plane, and / or change the focal length, and / or change the aperture, and / or adjust the focus to provide guidance to the target. 4. The CCTV system according to claim 1, characterized in that it further comprises a target detection sensor associated with the target recognition device, the target recognition device adapted to analyze a signal from the target detection sensor. 5. The video surveillance system according to claim 4, characterized in that the device selected from the group including infrared cameras, thermal imaging cameras, radar devices, a displacement sensor is used as a target detection sensor. 6. The video surveillance system according to claim 1, characterized in that the target recognition device is configured to calculate a velocity vector, the current coordinate of finding the target, and generating and transmitting to the mobile video camera the corresponding control signal to ensure that the mobile video camera is aimed at the target. 7. The video surveillance system according to claim 1, characterized in that the target recognition device is configured to transmit a control signal to the mobile video camera via a serial communication channel. 8. The video surveillance system according to claim 1, characterized in that the target recognition device is configured to determine the target by analyzing the signal from the video camera in accordance with parameters selected from the group including brightness, size, duration of stay in the controlled area, speed and direction of movement . 9. The video surveillance system according to claim 1, characterized in that a personal computer is used as a computing device with memory. 10. The CCTV system according to claim 1, characterized in that in the memory of the computing device using software tools a map model of the space controlled by the system is formed, made with the possibility of applying to it a target model that reflects the position of the target in the space controlled by the system, in accordance with the data obtained by processing in the target recognition device a signal received from a video camera, moreover, this card model is adapted for analysis by an operator. 11. The video surveillance system according to claim 10, characterized in that the card model is made three-dimensional. 12. A video surveillance method, comprising the steps of receiving a video signal of a controlled area of at least one video camera, transmitting it and analyzing it in at least one computing device with memory, characterized in that the signal received by the video camera is analyzed using a computing device using information with descriptions of alarm situations, pre-recorded in the form of a database in the memory of the computing device, according to the analysis results receive data about the target, form a control signal, transmit it to a mobile video camera, adapted to aim at the target in accordance with the control signal, using the mobile video camera receive a target image suitable for further analysis, and transfer it to the database for storage. 13. The method of video surveillance according to item 12, characterized in that according to the results of the analysis, an alarm is additionally generated and transmitted to the operator. 14. The video surveillance method according to claim 12, characterized in that the analysis is performed with the possibility of calculating the velocity vector, the current coordinate of finding the target, generating and transmitting to the mobile video camera the corresponding control signal to ensure that the mobile video camera is aimed at the target. 15. The video surveillance method according to 14, characterized in that the control signal is transmitted to the mobile video camera via a serial communication channel. 16. The method of video surveillance according to claim 12, characterized in that the analysis of the signal from the video camera is carried out taking into account the target parameters selected from the group including brightness, size, duration of stay in the controlled area, speed and direction of movement. 17. The method of video surveillance according to item 12, characterized in that a personal computer is used as a computing device with memory. 18. The video surveillance method according to claim 12, characterized in that in the memory of the computing device using software tools form a map model of the space controlled by the system with the possibility of applying to it a target model that reflects the position of the target in the space controlled by the system, in accordance with the data obtained when processing in the target recognition device a signal received from a video camera, moreover, this model card is adapted for analysis by an operator. 19. The method of video surveillance according to p, characterized in that the map model is formed three-dimensional.

Images