FR2855351A1 - Image authentication method for video surveillance system, involves comparing authentication code and pulsed luminance code, and generating signal to activate alarm if codes are different - Google Patents

Abstract

The method involves generating an authentication code (C) using a code generator (5), and applying the code (C) to an illuminator (2). A circuit extracts pulsed luminance code (C) from a video signal that is received from a camera (1). A logical comparator (8) compares the authentication and luminance codes, and generates a signal to activate an alarm, when the codes are different. The illuminator (2) controlled by the authentication code (C) transmits the pulsed luminance code (C) in certain points of an image that is captured by the camera (1). An independent claim is also included for an image authentication device.

Description

The present invention relates to an electronic authentication method

Image

  very high security, intended for video surveillance systems. The main purpose of this device is to warn of any attempt at fraud, by formally authenticating both the link and the image itself by acquiring data through the camera. The disappearance, alteration or displacement in time of said data, by occultation, displacement or substitution of their source validates an alarm. This feature allows a single device to perform simultaneously or as desired various complementary functions in the field of electronic security.

  Examination of the state of the art to date reveals, on the one hand, that almost all of the current video surveillance systems are devoid of devices allowing the security of the video signals between the cameras and the surveillance control unit and, on the other hand except that the few existing systems do not authenticate the image itself.

  Some multiplexed systems embed a number identifying the camera in the image. However, this identification number is most often generated by the multiplexer, either at the cable grouping point in the case of a long distance link, or at the control room and not in the camera itself, this which does not authenticate the end-to-end link. Furthermore, when this number is embedded in the image by the camera itself, it appears on a recording in static form and does not make it possible to ensure that the image reaches the control room in real time.

  Multiplexers are generally equipped with a device validating an alarm on the absence of N successive frames, drawing attention to a link interruption. Such devices do not make it possible to detect a substitution of the real signal by a recorded signal coming from a video recorder. This scenario, classic in cinema, is very easily achievable with current technological means. There are indeed quick connect assemblies designed to simplify connections over coaxial cables in corporate computer networks. This equipment makes it possible to very quickly carry out derivations on coaxial links. Substitution of the real signal by the simulation signal can be done without loss of frame by using a fast electronic switch triggered by the frame synchronization signal of the original video signal. Such a device, associated with a video recorder or an image box integrating a synchronous analog video camera and a synchronization regenerator slaved to the original signal, is capable of deceiving both an asynchronous link (synchronization signals generated by the camera) than a synchronous link (synchronization signals generated by the multiplexer). In the event of a rare high-end secure system, the cameras of which generate a code incorporated into the video signal, the substitution of the electronic signal by means of a recording would become difficult, if not impossible, if this code is of infinite length. and generated in real time.

  If electronic substitution becomes complex, on the other hand, nothing prevents an individual who has been able to access the base of the camera from presenting it with a photograph, a painted canvas, a backlit slide, a liquid crystal or cathode-ray screen (of preferably remanent), or any other display device. The majority of the 5 systems being sequenced at the control room to reduce the number of monitors, the installation of the decoy device could go unnoticed between two examinations of the camera concerned. Although this process is more risky to implement than the previous one for a high-flying burglar, it is conceivable and constitutes a share of risk to be taken into consideration.

  The purpose of the present invention is to remedy the security vulnerabilities of video surveillance systems, mentioned above. To this end, the method described below and shown in FIG. 1 is characterized by the fact that it formally authenticates the image, and by extension the video signal, and not the signal alone, by the acquisition of data through the camera (1).

  This authentication being carried out optically, by integrating into the field (D) 15 covered by the camera (1) at least one illuninator device (2) controlled by a code (C) which constitutes on the one hand the optical signature of the image and is also capable of transporting additional information. In addition to its very high degree of security, this particularity opens up possibilities for detecting presence in the field and transmitting information via the camera, which do not exist on conventional systems. 20 Note that the present invention is mainly dedicated to stationary cameras with fixed focal lenses. Mobile cameras allow any doubt to be raised at any time, by exploring the field. If however it is desired to authenticate permanently the signals coming from mobile cameras (particularly if their movement is automatic), it will be necessary to add to the system a device for tracking the illumination point, controlled by the motors of the camera or to provide the installation of several illuminators which will authenticate the image at intervals during breaks in predetermined sectors of the covered field. Unless the illuminator is placed exactly in the optical axis of the camera, the device does not accommodate lenses with variable focal length which induce, during their operation, a radial displacement of the point of illumination. The authentication of the image can only be effective in this case for a particular focal distance over which the objective must be systematically adjusted outside the observation ranges. It is possible to provide the device with a timing circuit which automatically returns the lens to the appropriate setting after a predetermined time. Whatever type of camera is used, it will always be possible to perform a point-in-time authentication in order to remove doubts, in the event of an image jump suggesting a fraud attempt for example. The pointer will then be positioned manually on the illumination point by the operator. A manual control (3) is provided for this purpose. An audible (11) and (or) luminous signal will confirm or invalidate the presence of the valid code in the image.

  The method according to the invention performs formal authentication of the image by placing in the zone protected by the camera (1), and in the field (D) thereof, one or more illuminating devices (2) receiving at distance by cables (wired, optical fibers, alternating network by high frequency on carrier currents) or wireless (by hertzian way), a falsifiable code (C) generated by a circuit (5) contained in an electronic unit (B) located at level of the control room and which contains all the circuits necessary for the implementation of the device. If this proves necessary, and in the event that none of the 10 aforementioned modes of transport of the code could be implemented for technical reasons or site configuration, it remains possible to route the code to the room to be monitored using the video coaxial cable itself. This possibility which is a priori only offered by synchronous cameras, that is to say the synchronization signals of which are supplied by the multiplexer (X) however requires the use of additional means consisting on the one hand of a device (12) located in the electronic unit (B) allowing the modulation in all or nothing or in width and at the rate of the code (C) of the line or (and) frame synchronization pulses, and on the other hand d '' a device (13) located at the output of the camera (1), intended to restore the modified pulses, for the proper functioning of the camera, and to restore the authentication code (C) to make it available to ( of) illuminating device (s). The code (C) is synchronized by the synchronization signals extracted from the video signal by a separator circuit (6) which provides all the useful signals such as line synchronization, frame synchronization, even or odd line information. Respecting the unity of invention, the very nature of the code is left to the discretion of the manufacturer. Let us simply point out that only a permanent flow code generated in real time makes it possible to achieve a high degree of security. The prototype system of the present invention implements a tamper-resistant proprietary code which has been the subject of a separate filing.

  Said code (C) is applied to the emitting component of the illuminator (2) (light-emitting diode or other component of compatible characteristics.

  Said illuminators (2) are preferably not very directional so as to be clearly visible whatever the relative position of the camera. They may be provided with a diffusing lens, preferably of a dark color and not reflecting, in order to increase the contrast from the point of view of the camera. They transmit at certain precise points of the image, judiciously chosen (in the form of a spot visible on the screen) a pulsed light code (in the visible or invisible spectrum depending on the specific characteristics of the cameras). This light code is extracted from the video signal by an appropriate circuit (7) which analyzes the luminance or chrominance characteristics of the signal at the time of scanning of the illumination point, the coordinates of which (abscissa and ordinate) are designated in l image by the pointing device (4), which is essentially composed of a line counter and a frame counter slaved to the synchronization signals supplied by the separator circuit (6). The code (C ') extracted from the analysis point is subjected to a logical comparator (8).

  Said comparator circuit (8) compares the code (C) transmitted with the code (C ') extracted from the video signal and generates a signal which validates an alarm in the circuit (9) in the event of a difference observed between the codes C and C'. This alarm (A), materialized by a logic level, is capable of actuating a relay (10). The contacts of said relay (10) make it possible to trigger any useful device, such as the audible warning (11) or the motor of the filter holder mount (M). The multiplexers of the video surveillance control room generally include forcing inputs intended to receive contacts from presence detection systems. The relay contacts (10) are available for such use and allow viewing of the priority image. The device remaining active 15 even when the image is not observed or even selected.

  The device as described comprises a complementary security system which consists in doubling the analysis circuits in order to scan at least one neutral point of the image, in addition to the point of illumination. For simplification, these complementary circuits, identical in principle to the circuit which analyzes the point of illumination, are not shown in the drawings. This neutral point is preferably chosen on a dark area of the image, close to the light spot produced by the illuminator. The detection at neutral point of a code identical to the transmitted code, or identical to the code received from the illumination point, reveals an attempt to dazzle the camera by a decoy illuminator taking the code from the original illuminator. Such an attempt generates an alarm (9). This alarm information can be delivered in the form of a logic level (A) and / or on the contacts of a relay (10), as in the case of a code error of the original illuminator .

  The type of sensor fitted to the camera (I) and the spectral response thereof will guide the choice of means to be used to extract the data from the video signal. For a monochrome camera or a color camera with good infrared sensitivity, the information will be extracted from the luminance level. A color camera not very sensitive to infrared will involve the use of a visible light illuminator whose color will be chosen so as to provide the best contrast depending on the color temperature of the lights or the dominant color of the back plan. The analysis circuit will then extract the code of the appropriate channel from a discriminator circuit providing the chrominance signals. - 5

  The implementation device as described is mainly designed to complement a video surveillance system. To this end, said device is equipped with all the functionalities necessary for implementing the method autonomously and entirely electronically. Said device is adaptable to any analog or digital system carrying composite video. The majority of video surveillance installations in service use this technology for the acquisition and transport of video. The same is not true for image processing, which for reasons of display, operating and storage convenience is increasingly being handled by a computer-based digital unit. While responding to the same principle 10 of acquiring an authentication code by optical means, the method can then see its hardware implementation device simplified. In a variant, said device can in fact be reduced to the signal separator circuit (6) and to the code generator (5) synchronized by said separator circuit (6). The rest of the processing will then be carried out by software, the reference code (C) coming from the circuit (5) being supplied to the computer system 15 by a simple logical interface. The advantage of such digital processing lies in the possibility of examining an unlimited number of neutral points or illuminators without multiplying the analysis circuits, thus combining on a pre-existing computerized system, performance and economy of means.

  We will describe below some of the many advantages and applications which arise from the process which is the subject of the present invention. While referring to the same implementation device, these applications constitute variants in use.

  - The process is economical. Its implementation device is technologically simple. The method excludes the use of an expensive 25 frame memory. The system scans only a few points of the image requiring only the processing in real time of a few isolated bits, without storing the image. In its basic fully electronic version dedicated to analog installations, it does not use IT and does not require additional IT equipment.

  - Easy to install and use, the device for implementing the method is directly compatible with all analog video surveillance equipment on the market transporting composite video.

  - The device is capable of operating as a presence detector by occultation, extinction or displacement of the illuminator (2), or even as a fire detector. In fact, the virtual line (L) connecting the illuminator (2) to the camera (1) behaves like an optical light curtain. The interruption of said barrier by an obstacle (body, solid, smoke or fog) validating an alarm. The presence of smoke is also able to broaden by diffraction or diffusion the light spot seen by the camera. This widening can reach a neutral point judiciously positioned and validate an alarm. This barrier can cover a considerable distance provided that time delays are taken into account by appropriate means. The efficiency of the device outdoors depends on the emission power of the illuminator (2), since the optical link (L) is in this case dependent on atmospheric conditions, in particular in the event of precipitation or dense fog. The interposition of a mobile infrared filter (F) on a motorized mount (M), stopping the visible spectrum, makes it possible to reinforce the contrast and make the link more reliable in barrier mode. This arrangement allows at any time and particularly during an alarm, to go into observation mode by removing the infrared filter (F) placed in front of the objective. The main advantages of using a video camera as a receiver, in a optical barrier, resides in the simplification of the emission optics which does not require the generation of a very narrow beam, and the ease of alignment which is carried out visually, precisely and quickly, by superimposing on a video screen a pointer to the illumination point by means of 15 manual controls (3) accessible remotely, without mechanical manipulation.

  - Incorporated or connected to any presence detection device (DP) (door contact, microwave radar, passive infrared detector, etc.), the illuminator (2) allows when its emission is interrupted by said detector at by means of an optional controlled contact (K), coming from a relay or other switching device, to draw attention to the point observed by the camera. It will thus be possible to protect the camera itself by an appropriate detector directly monitoring it (radar, etc ...) or monitoring an access close to it (door, corridor, etc ...). In addition, fixing the illuminator (2) to a door saves a sensor. The movement of said illuminator (2) in the field (D) of the camera validating an alarm 25 when the door is opened, by disappearance of the code received.

  - In conjunction with the surveillance of the premises themselves, the protection of precious objects in museums is a particularly interesting application of the process.

  One could, for this particular type of application, consider using a directional illuminator (laser or collimated beam) whose beam would be pointed at the surface of the object 30 to be protected. The analysis point processed via the camera would therefore be a clear spot on the surface of the object. We will call the virtual spotlight the spot of light produced by a directional illuminator. The adoption of an infrared laser would make this protection unsuspected and any movement of the object would cause an alarm, as well as the interception of the beam of the illuminator. Although attractive at first glance, the implementation of this method is subject to conditions. It is indeed possible for an individual equipped with an infrared vision system, to interpose in the beam a screen close to the surface of the object. It is then this screen which would carry the stain and no longer the protected object. It is also excluded to thus protect a painting hanging on the wall, because the removal of it would not cause a significant displacement of the light spot.

  The use of this method is to be reserved for three-dimensional objects placed at a sufficient distance from the walls and far from the center of the image to benefit from a significant displacement of the spot due to the parallax. We most often prefer to adopt a diffuse illuminator, well suited to the protection of three-dimensional objects, such as sculptures or statues, provided that they have a perforated part which can let see through it the light spot produced by the illuminator diffuse. It will not be possible to remove the object without briefly concealing this spot. It is still possible to implement another variant combining the two methods. in this variant using a laser, the light spot is projected onto a catadioptric reflector which appears as a virtual illuminator and restores its properties.

  In this case it is the laser beam which passes through an orifice of the object, the optical axis camera-reflector passing outside said object. The movement of the object causes the laser to be obscured and the catadioptric reflector to be turned off from the point of view of the camera.

  It should be noted that the few drawbacks mentioned present only a tiny risk if human permanence is ensured under management. Indeed, the installation of the hardware elements necessary for the fraud of the system, such as mirrors or screens 20 requires great precision and a long time of presence in the field of the camera, time much greater than the period of scrutiny of n any sequencer. For an unsupervised system, entrusting the process with the full-fledged alarm function, the risk is scarcely higher since the implementation of the evoked means of fraud is extremely hazardous given the high sensitivity of the system.

  - It is possible to independently process the even and odd frames, and to assign them different functions. The separator circuit (6) provides the necessary frame identification signals for this purpose. It will thus be possible to assign for example the even frame to the processing of the authentication information of the illuminators (2) dedicated to authentication, while reserving the odd frame for the acquisition of service signals 30 from the illuminators auxiliaries assigned to the transmission of specific data. The possibility offered by the process, of acquiring information other than the authentication code, will make it possible by way of illustration, to transmit to the control center signals coming from an access control keyboard via the camera. To this end, said keyboard incorporating an illuminator will be placed in the field of the camera. An individual appearing in front of the keyboard will intercept the optical beam (virtual beam because the illuminator being designed with little direction to facilitate its installation, it is perceived as a beam from the point of view of the camera alone) which will attract an alarm l attention of the control room on the camera concerned. As long as the optical beam is interrupted, that is to say as long as the keyboard illuminator is not visible to the camera, said keyboard cannot be used by an unauthorized person. An authorized person will set aside to compose their code, which will be transmitted to the control room via the camera during the even or 5 odd frame assigned to this treatment. The code assigned to access control can be derived from the authentication code, which will also make it falsifiable, since it is generated in real time.

  One can also imagine a discreet round-pointing device in the form of a portable illuminator that the security personnel will momentarily place at a precise point in the field of each camera during the premises inspection tours. Such a device dispenses with the installation of specific readers dedicated to pointing, as soon as the premises are under video surveillance. A simple plate fixed to the wall allows the placement of a portable illuminator, and constitutes an excellent solution against vandalism, because no vulnerable element such as a reader, is installed permanently.

Claims (10)

  1) Electronic video authentication method for general-purpose use, simultaneously securing the video link (V) between the camera (1) and the control room, and authenticating the image itself, in real time, by the acquisition of data through the camera (1), said method being characterized in that this authentication 5 is carried out optically, by integrating into the field (D) covered by the camera (1) at least one illuminating device real (2) or virtual (the projection of a beam on a surface, considered as a virtual illuminator), controlled by a code (C) which constitutes the optical signature of the image and which is capable of transporting additional information, said code, generated (5) in an electronic box (B) and routed to the illuminator (2) 10 by ordinary (wired, fiber optic, carrier currents, wireless) or specialized (devices 12 and 13) being received in light form through e the optics of the camera, said code being simultaneously extracted (C ′) from the video signal at the point of illumination designated by a pointing device (4) slaved to a manual control (3) to be compared by hardware (8 ) or software with the generated code (C) considered as a reference, the result of this comparison generating an alarm (9) materialized by the opening or closing of a contact (10) usable for various purposes, the method also making it possible, using at least one additional analysis circuit identical in principle to the first, by pointing and analyzing a neutral point of the image, to detect an attempted fraud by dazzling the camera, the method allowing to simultaneously fulfill various functions in the security field, such as optical barrier, presence detection, object protection, round-off pointing, data acquisition.   2) Device for implementing the method according to claim 1 characterized in that it comprises means (5) for generating in real time a signature code (C) intended for one or more illuminating devices ( 2) placed in the field (D) of the camera (1).   3) Device according to claim 2, characterized in that it comprises means for transmitting the authentication code (C) in the form of a light spot visible on the screen, said illuminator (2) being used in direct connection or by reflection on a surface and emitting in a spectrum compatible with the specific characteristics of the camera used, but preferably invisible to the human eye (infrared, ultraviolet).   4) Device according to claims 2 and 3 characterized in that it comprises means (12 and 13) allowing by modulation and reproduction of the synchronization signals of the video signal, to route the signature code (C) to the ( the) illuminating device (s) by means of the cable serving the camera (1) when for technical reasons or site configuration, none of the ordinary solutions (sniffing, optical fiber, carrier currents, hertzian) is possible .   5) Device according to claims 2 and 3 characterized in that it comprises means (4) for generating on demand on the screen of a video monitor or a television, a pointer (reticle or any other visible artifact) used to designate the precise position of the illuminating device (s) and of the neutral point (s), the position of said pointer being adjustable by means of a manual control (3 ). 10 6) Device according to claims 2, 3 and 5 characterized in that it comprises means (7) for extracting from the video signal and therefore from the image, the authentication code (C ') perceived on the screen as a bright spot and designated by the pointing device (4).   7) Device according to claims 2 and 6, characterized in that it comprises means (8) for comparing the code received (C ') through the camera to the code (C) sent to the illuminator (to the illuminators ) (2), said device validating an alarm (9) according to the result of this comparison. 20 8) Device according to claims 2 and 5, characterized in that it comprises the means of detecting an attempt to dazzle the camera (1) by an illuminator foreign to the system by analyzing at least one neutral point of the image.   9) Device according to claims 2 to 8, characterized in that it comprises means for transmitting in light form via the camera, and by the channel of the authentication code, useful data, widening the field of use d '' a single device for various complementary and simultaneous applications in the security field.   10) Device according to claim 2 characterized in that it comprises software means for extracting the code (C ') from the image in its digital form as well as hardware means for interfacing the code generator (5) with a computer-type digital system.