FR2766312A1 - PROCESS FOR TRANSMISSION OF INFORMATION BETWEEN A RECEIVER, IN PARTICULAR A TELEVISION RECEIVER, A BOX AND A VALIDATION STATION, IN PARTICULAR A VOICE SERVER, AND CORRESPONDING BOX - Google Patents

Abstract

The invention concerns a method which consists in integrating in the broadcast transmission main data corresponding to a specific event occurring during the programme, and an identifier of the event and an additional indication for defining a time frame for broadcasting the event; storing in the validating station (SV) each event identifier. After reception in the case unit (BT), storing therein each specific event identifier having actuated at least once in the case unit the control element (TF) in the time frame defined by the additional indication associated with this event; restoring outside the case unit and to the validating station, the case unit identifier and the event identifiers for analysis.

Description

Method for transmitting information between a receiver, in particular a television receiver, a box and a validation station, in particular a voice server, and corresponding box.

 The invention relates to the interactive transmission of information between a receiver reproducing a radio broadcast, a box having at least one manually operable control element such as a key, and a validation station.

 The invention finds a particularly advantageous but nonlimiting application in the interactive connection between the host of a television program offering objects for sale, a potential buyer discovering these objects on his television set, and a voice server. intended to record potential viewer commands.

 The invention also applies to the interactive connection during a program containing a flood of advertising messages, between advertisers, a viewer discovering these messages on his television set and interested in some of them and a voice server intended for example to provide more information on the products presented on television.

 No such interactive information transmission systems are known at present. The invention aims to propose.

 Furthermore, these systems pose numerous technical problems. Indeed, the box in possession of a user, must be both simple to use on the technical level, of an industrially reasonable cost, even limited, and be able to contain all the information and characteristics relating to the objects presented on the receiver while limiting the size of the memory contained in this object.

 In addition, one of the characteristics of the invention consists in ensuring rigorous time synchronism, in particular between the box and the receiver reproducing the transmission, so as to ensure the reliability of the interactive system by avoiding in particular errors between the command. performed by the viewer in view of the objects on his receiver and the object which he will ultimately receive from a command validation station, to which he is not physically connected, and distinct from the transmitting station proposing said objects .

 This rigorous time synchronization must be carried out while avoiding installing, in particular in the portable object possessed by the viewer, time counting means that are too complex to implement, and therefore too costly.

 The interactive method of transmitting information according to the invention provides a solution to all these problems.

 More precisely, according to the method according to the invention, at least one group of specific information intended to be restored in a predetermined time period is incorporated into the radio broadcast, this group of specific information comprising a sequence of main information corresponding to a specific event occurring during the broadcast.

 In particular, when the radio broadcast is a television broadcast, each specific event may consist in filming for a given period of time a specific object intended to be possibly purchased by a holder of the box who watches the broadcast on his television set. Each sequence of main information restored then advantageously consists of the image of the corresponding object on the screen of the television receiver. The time range associated with each image can thus be delimited temporally with respect to two temporally spaced shots of the corresponding object.

 In addition to this sequence of main information, the group of specific information intended to be restored in the predetermined time range comprises reference information containing at least one identifier of the event, for example an identifier of the object, and in practice also an identifier of the program for example, as well as an auxiliary indication making it possible to define said time range.

This auxiliary indication can very simply be the duration separating the two shots.

 When the radio broadcast is a television broadcast broadcasting a stream of advertising messages, one of these advertising messages may consist of said specific event, the main information sequence then corresponding to the images of the advertising message broadcast on the television receiver. . The auxiliary indication can then quite simply be the duration of the time period of broadcasting of the corresponding advertising message.

 According to the invention, there is also stored in the validation station, for example a voice server, so-called validation information containing at least each event identifier and possibly additional information on the event, for example the object.

When the broadcast broadcast is a delayed broadcast, this information can be stored in the memory of the server before the broadcast of the broadcast on the television set, for example by an appropriate computer network. When it is a live broadcast, the various event identifiers can be sent in real time, also via a computer network, to the voice server.

 With regard to the portable box, which is equipped with at least one manually operable control element such as for example a key, each reference information is received at the level of the box, for example by means of a specific sound signal such as 'an acoustic click or "Jingle" in English, and a storage indication is stored in the box containing at least the identifier of each specific event which gave rise to a detection within the box of at least one actuation of the control element in the time range defined by the auxiliary indication associated with this event.

 Concretely, in an application of the "tele-shopping" type, the owner of the box will activate the control element each time that he is interested in the acquisition of an object whose image appears on his television receiver. . Restitution outside the box and to the validation station, restitution information containing an identifier of the box and said storage indication.

 With regard to the validation station, the restitution information received and the validation information stored are analyzed within it, so as to possibly associate at least one event identifier with the identifier of housing. This analysis processing will therefore generally constitute a comparison of the event identifiers received from the box with those stored in the memory of the server. In the event of a match, there will then be a correspondence between the identifier of the event, i.e. for example the identifier of the object, and the identifier of the box from which the address of the viewer may be known. The corresponding object can then be returned to its recipient.

 The invention is therefore remarkable in that it avoids storing in the box all the characteristics of the objects that the viewer wishes to acquire, as well as all the temporal characteristics of the appearance of these objects on the screen. Some of this information is transferred to the server memory, which considerably reduces the memory size of the object's memory. Furthermore, the fact of transmitting to the box an auxiliary indication making it possible to define each time range and to detect any actuation of the control element of the box in this time range makes it possible to dispense with installing in the object a clock indicating the universal time, with all the problems of temporal drift which are grafted there, to ensure in a simple way a rigorous temporal synchronism between the broadcast broadcast and the box and the server, to free itself from storing in the server of time references relating to the occurrence of the event on the receiver and to transmit only an identifier to the server, thereby minimizing the risk of errors.

 Furthermore, within the meaning of the present invention, the term "broadcast" applies to the broadcast of sound (for example intended to be received on a radio) and / or television information, or even more generally to any information broadcast by radiofrequency waves.

 In order to further minimize the risk of error, it is advantageous to provide mutually temporally spaced time ranges.

 Although each reference information can be restored at any time in the time range, the auxiliary indication can then for example include the time that has already elapsed since the start of the time range as well as the time remaining until the end of the time range, it is particularly advantageous that each reference information is restored at the end of the corresponding time range. This reduces the risk for the user of the box to miss reception of the reference information, for example if he is not present in front of his television set with his box in hand at the start of the message. advertising for example.

 According to one embodiment of the invention, the detection of an actuation of the control element of the housing within a time range, comprises on the one hand the initialization of a time counter at a predetermined initial value, for example the value zero, and the analysis, upon reception of the corresponding auxiliary indication, of the current value of the counter with respect to said time range taking into account said auxiliary indication.

 The use of a counter that can be initialized on each actuation of the control element of the box and the use of its current value when the auxiliary indication is received, makes it possible, in an extremely simple manner, to determine whether the actuation is in the time range. This notably avoids the use of a very precise real time clock, which is therefore very expensive, and even allows the use of a time counter, which is not necessarily very precise, and which can therefore exhibit a certain temporal drift. In addition, such an implementation mode offers the user the possibility of actuating the control element of its housing several times in succession without disturbing the final analysis carried out within the housing since, of all ways, in this case, only the last actuation will be counted.

 In addition, in combination with reference information restored after actuation of the counter, for example at the end of the time range, such an implementation mode offers the advantage of great technical simplicity since it does not require than reading an incrementing or decrementing counter.

 In a very simplified variant of the invention, the reference information can be reproduced orally by the host of the radio broadcast. The latter can, for example, invite the owner of the box to actuate a specific key on the keyboard, then corresponding to the reception of the auxiliary indication, in the case for example where the duration of the time period is predetermined and stored in the case memory. Similarly, the identifier of the object can be indicated orally by the facilitator by means of a code that the user then dials on the keyboard of his box.

 That said, the reference information can also be restored in the form of a specific optical signal restored by the receiver in the direction of appropriate means of the housing.

 Better, the receiver can reproduce each reference information within a specific sound signal, for example an acoustic "jingle". We then receive the specific sound signal at the level of the box and extract said reference information therefrom.

 Such an implementation variant thus allows transparency in the transmission of reference information vis-à-vis the user of the box and is particularly useful when transmitting advertising messages.

 The reference information is generally a digital message made up of several bits. Conventional coding systems for a digital message in a sound signal generally use specific frequencies of the acoustic signal for the transmission of the message. However, the reflection of the acoustic signal restored and containing the message, on the walls of a room for example, generally results in a modification of the frequency spectrum of the signal, which then poses problems for the correct decoding of the messages contained in the signal. acoustic received.

 According to a particularly advantageous embodiment of the invention, a first elementary acoustic signal having predetermined characteristics is then produced, and a second elementary acoustic signal reversed in phase with respect to the first acoustic signal, the two elementary signals having durations equal elementaries.

Any bit having a first logical value is then coded by a pair of first consecutive elementary acoustic signals or by a pair of second consecutive elementary acoustic signals. Any bit having a second logical value is then coded by a pair of different consecutive elementary acoustic signals. The specific sound signal is then developed from the first and second elementary acoustic signals which define said reference information.

In addition, an additional sound signal can be incorporated. And this specific sound signal intended to be broadcast and then reproduced by the receiver is incorporated into said broadcast.

 In other words, the logical value of each bit of the reference information is coded directly from the sole form or temporal evolution of the elementary signals without it being necessary to perform any processing on these elementary signals, frequency for example, to obtain the coding of the bit. In other words, the information to be transmitted is then not contained in the particular frequencies of the specific sound signal but in the succession of two consecutive sound elements which directly define the state of the bit to be transmitted, for example according to whether 'they are the same or different.

 The additional sound signal which can optionally be incorporated into the specific sound signal, is for example background music. This added signal has no functional role in the actual coding of the bits of the restitution information and must be chosen so as to remain at a level much lower than that of the coded signal containing the useful information. As an indication, this additional signal should not exceed in amplitude 10% of the coded signal. This additional signal therefore only serves to make the tones more pleasant to the ear.

 To decode the logical value of a bit transmitted by such a specific sound signal, an autocorrelation of the specific sound signal received is advantageously carried out over a duration at most equal to the elementary duration of the two elementary signals used, the result of this autocorrelation making it possible to deduce the logical value of the bit received.

 Although it is preferable to perform the autocorrelation of the signal over a duration equal to the elementary duration, it is not impossible to perform this autocorrelation over a shorter duration.

 Of course, for the decoding of the reference information which comprises several bits, a succession of autocorrelations of the specific sound signal received is carried out.

 Concretely, the specific sound signal is reproduced for example by means of the loudspeaker of the television set or the radio set and this specific sound signal is then advantageously captured at the level of the housing by means of an acoustic sensor, for example a microphone, so as to decode the bits of the reference information. Thus, in this case there is no wired link between the television receiver and the box comprising the microphone.

 Such an implementation of coding and decoding of bits makes it possible to overcome as much as possible the problems of reflection of acoustic signals.

 As regards the restitution information intended for the validation station, it can advantageously be restored outside the housing in the form of an acoustic signal, for example an acoustic barcode which is transmitted to the voice server by through a telephone link.

 The invention also relates to a housing, usable for interactive transmission of information between a receiver reproducing a radio broadcast, the housing and a validation station. This box includes means for receiving reference information capable of being reproduced by the receiver and containing at least one identifier of a specific event broadcast during transmission and an auxiliary indication making it possible to define an associated broadcasting time range. at this event. The housing also includes a first memory containing a housing identifier, as well as a control element, for example a key, capable of being actuated manually by a user of the housing.

 There are also provided detection means capable of detecting any actuation of the control element and processing means suitable, in the presence of at least one actuation detected in said time range, for storing in a second memory a storage indication containing at least the identifier of the specific event contained in said reference information. The housing also includes controllable restitution means for restoring, outside the housing, restitution information containing the housing identifier and the storage indication.

 According to one embodiment of the invention, the detection means comprise a time counter which can be initialized on command to a predetermined initial value. There are also provided first control means capable, in the presence of any indication of actuation provided by the detection means, to initialize the counter at said predetermined initial value, second control means capable of reading the current value of the counter during reception of said reference information, and control means capable of determining the possible presence of an actuation in said time range as a function of said initial value, said current value and said auxiliary indication.

 The detection means can comprise a third memory as well as second control means capable of storing in the third memory said actuation indication, for example the logic value 1, on each actuation of the control element.

 In the case where the reference information is reproduced within a specific sound signal, the means for receiving the box comprise an interface for receiving said specific sound signal, as well as decoding means capable of extracting said information from the sound signal. reference. These decoding means thus advantageously include autocorrelation means capable of performing autocorrelation of the specific sound signal received over a duration at most equal to the elementary duration of the elementary signals used to code the bits of the reference information.

 Furthermore, when two elementary acoustic signals reversed in phase are used, the decoding means advantageously comprise means for shaping the specific sound signal received comprising means for comparing this signal with a reference value so as to obtain a binary comparison signal.

 I1 is also provided means for sampling the binary comparison signal and two cascaded registers for storing the two groups of samples thus obtained. The autocorrelation means preferably comprise means of comparison of the EXCLUSIVE OR type capable of carrying out a sample-to-sample comparison of the two groups of samples, as well as an auxiliary counter capable of recording the number of samples of identical value or different, and an output flip-flop able to take the logic value 0 or 1 as a function of the comparison of the counting value with respect to two predetermined thresholds.

 These various means can for example be produced by means of a specific integrated circuit (ASIC).

 The decoding means can also further comprise a microprocessor capable of reading the content of the output flip-flop so as to deliver the bits of the reference information.

 While various data transmission networks can be envisaged for the link between the portable box and the voice server, such as for example fiber optic transmission networks using appropriate call stations, it has been considered currently preferable, in particular for use existing infrastructure and for cost reasons, use a telephone network.

 Given this particular choice, it would be possible to use for example the conventional mode of telephone transmission known by a person skilled in the art under the acronym FSK (Frequency Shift Keying) which uses two transmission frequencies to transmit a computer bit according to its logical value 0 or 1. However, it has been observed that it was preferable to use, instead of in particular a transmission mode of the FSK type, coding using a single-frequency signal. This avoids disturbing the transmission of data with parasitic frequencies and can also lead to a hardware simplification of the transmitter.

More precisely, said restitution information being made up of a plurality of bits, the restitution means advantageously comprise:
an activation element, for example a push button or even one of the keys of the keyboard mentioned above,
coding means, for example produced in software within a microcontroller, capable, in response to actuation of the activation element, for delivering, for one bit of said restitution information having the logic value "1", a logic signal having a high state for a first predetermined duration and a low state for a second predetermined duration. For a bit of said restitution information having the logic value "0", the coding means deliver a logic signal having the low state during all of the first and second durations, and
- an electro-acoustic transducer, for example a piezoelectric blade, delivering no acoustic signal in response to a loqique signal in the low state, and delivering in response to a logic signal in the high state, a mono acoustic signal -frequency whose frequency is located substantially in the center of the bandwidth of a telephone line.

 Such an acoustic coding system is similar to an "acoustic barcode".

Other advantages and characteristics of the invention will appear on examining the description of a non-limiting embodiment and the attached drawings, in which
FIG. 1 is a very schematic block diagram of an interactive information transmission system according to the invention,
FIG. 2 illustrates in more detail but still schematically the internal architecture of an autonomous portable housing according to the invention,
FIG. 3 schematically illustrates an operation of the system according to the invention,
FIGS. 4 and 5 illustrate a restitution of information using an acoustic bar code according to the invention,
FIG. 6 schematically illustrates two elementary acoustic signals used in the invention for coding the reference information,
FIGS. 7 and 8 illustrate two other examples of elementary acoustic signals,
FIG. 9 illustrates two examples of coding a bit of logical value 1 contained in reference information,
FIG. 10 illustrates two examples of coding a bit of logical value 0 contained in reference information,
FIG. 11 illustrates the coding of a digital message comprising several bits and containing reference information,
FIG. 12 more particularly illustrates the internal structure of a transmitted digital message,
FIG. 13 is a schematic block diagram of a transmission device and of a device for decoding a specific sound signal containing reference information according to the invention, and,
FIGS. 14 and 15 illustrate in more detail schematically certain parts of the decoding device of FIG. 13.

 An embodiment of the invention will now be described in more detail, more particularly focused on an application of the "teleshopping" type, although, as will be seen below, the invention also applies to interactivity between a user. of the housing and an advertiser via a television program reproducing a stream of advertising messages.

As illustrated in FIG. 1, the interactive system SYS comprises an emission station STM capable of broadcasting an ERD emission, in particular a television emission, produced on a television set very schematically represented under the reference
PTV. On this television set, there are, in particular, CM cameras intended to film objects, here two, OB1 and OB2, capable of being acquired by a viewer watching the ERD program on his television.
RCP.

 A validation station SV, for example a voice server, is connected to the television set PTV via a data transmission network RTD, in particular a conventional computer data transmission network.

More specifically, the SV voice server has an interface
INT 1 connected to the RTD data transmission network. The voice server is also articulated around MG management means which can be incorporated in software within a computer. These management means interact with a memory station BD such as a database, as well as with a second interface INT2 connected to a telephone network, such as the switched telephone network PSTN.

 On this PSTN switched telephone network, a telephone set is also connected, the CT handset of which is able to cooperate acoustically with an electro-acoustic transducer, such as a piezoelectric blade BZ, of a portable portable unit BT in particular equipped on one of its faces of a CL keyboard.

 As illustrated very diagrammatically in FIG. 2, the hardware architecture of the BT box is articulated around a microcontroller PR, for example 4 bits, incorporating in software the different functional means of the box. However, it is not excluded to provide a hardware implementation using a specific integrated circuit (ASIC).

 A first memory MM 1, for example a read-only memory, containing an identifier 1DB of the box, for example its serial number, is connected to this microcontroller, a second memory MM2, also a read-only memory, intended to store at least one indication storage containing at least one identifier ID1 of an object for which the user has expressed an interest. However, in practice, this second memory MM2 can also store a 1DM identifier of the program in question, and possibly, in particular in an application of the advertising type, an identifier of the television channel and / or an identifier of the advertiser. .

 The keyboard CL includes a specific key TF, forming the control element of the box, and intended to be pressed by the user when the latter is interested in the object appearing on the screen. Also, there is provided a third memory MM3, for example a random access memory, intended to store on a bit, a logic value representative of an actuation of the corresponding key of the keyboard. Thus, for example, the value 1 will be representative of an actuation, while the value 0 will be representative of an absence of actuation.

The PR microcontroller is also connected, in addition to the keyboard
CL, a CMT time counter, an AFF liquid crystal display, a BZ transducer, typically a piezoelectric blade, as well as means for receiving a specific sound signal, for example an acoustic jingle, comprising a micro MC associated with means MRS which will be discussed in more detail on the structure and functionality below.

 All of these elements are supplied from ALM supply means, such as batteries. The microcontroller also incorporates in particular the software for managing the operation of the keyboard.

 We will now describe the operation of the system according to the invention with particular reference to Figure 3, and in the context of a particular application of the type "teleshopping".

 It is also assumed that the television program is broadcast on a delayed basis.

 On the top of Figure 3, there is shown the time diagram representative of the progress of the broadcast. This starts at DE time and ends at FE time.

 For example, after having briefly presented the characteristics of the various objects that can be acquired by viewers, the host will specifically invite viewers to order or not the first OB 1 object. In this regard, the CM camera will then fix the object OB1 during a time range PLT1 starting at time TP1 and ending at time TP2. The image IM1 of the object is then restored throughout the duration of the time range PLT1 on the screen of RCP receivers.

 The same operation is performed for the OB2 object. However, in order to minimize the risk of error as much as possible, the time range PLT2 associated with the object OB2 and delimited by the time instants TP3 and TP4 is temporally spaced from the range PLT1. During this entire time range PLT2, the image IM2 of the object OB2 is reproduced on the screen of the receivers.

 In addition, preferably at the end of each time range, a specific sound signal JJ 1 (JJ2) is reproduced by the receiver and contains reference information including, in the example described, an identifier ID1 (ID2) of the object under consideration, for example a reference code, as well as an auxiliary indication enabling the corresponding time range to be defined. This auxiliary indication may be the actual duration of the time range or else a duration number corresponding to a number prerecorded in a memory of the unit and corresponding to a predefined duration.

In addition to the object identifiers, an IDM identifier for the program, for example the current date and the game number, can also be contained in the reference information transmitted by the sound signals.
JJ1 and JJ2.

 The identifiers ID1, ID2 of the various objects as well as the identifier of the program, are moreover transmitted to the voice server for the purpose of storage in the database BD.

 In general, each press by the user of the box, on the key TF of the keyboard CL, leads the microcontroller PR to deliver an initialization signal to the counter CMT so as to initialize it at an initial value, for example 0. At the same time, the processor PR stores in the memory MM3, the logic value 1 corresponding to an actuation.

 From this moment, the CMT counter counts the seconds. If the TF key is pressed again, the CMT counter is again reset to 0.

 When the television program is broadcast, the viewer, equipped with its box, switches it on by means of a specific button. As a variant, provision may be made for the housing to be started up automatically as soon as the TF key on the CL keyboard is pressed.

 When the image IM1 appears on the screen of his television, the viewer, who is supposed to be interested in the object presented, actuates the F key, for example at time T10. The CMT counter then counts or counts down at the rate of one second. Logical value 1 (actuation indication) is stored in memory MM3. When receiving the reference information contained in the specific sound signal JJ1, and intervening at time T 1, the processor PR checks the value of the content of the memory MM3 and, if this is representative of an actuation of the TF key reads the current value of the counter CMT, which allows it to determine the time interval ATl that has elapsed since said actuation of the TF key. The processor PR then compares the time interval AT1 (or directly the current value of the counter if the initial value of the latter is zero) with the auxiliary indication contained in the reference information transmitted by the sound signal JJ1, of so as to determine whether said actuation of the TF key took place within the range PLTî. If this is the case, the microcontroller PR stores in the memory MM2 the identifier ID1 of the object, and, optionally, the identifier IDM of the transmission.

 It is also possible, which is particularly advantageous in a flood of advertising messages application, to equip the housing with an additional time counter (not shown in the figures for simplification purposes) delivering a time indication representative of the current time. It is thus possible to use a counter counting down or counting down at the rate of one second from a predetermined initial value representative of a predefined initial date. In this case, the microcontroller PR will also, by pressing the key TF, record the value of the auxiliary counter, (corresponding to time T 10) and also store it in the memory MM2 in correspondence with the indications previously stored.

 Each time it is stored in the MM2 memory, the microcontroller PR increments the display AFF. Indeed, it is advantageously provided that the memory MM2 can be broken down into several memory areas each intended to receive the identifier of a particular object. The indication provided by the display then allows the user to know the filling rate of his memory.

 If the viewer is also interested in the object OB2, he repeats the same operations, at time T20 for example, which is located within the time range PLT2.

 It should be noted here that the invention makes it possible to manage untimely presses of the TF key by the user, or else to take account of presses which do not correspond to a displayed object.

Indeed, assuming that the viewer presses the button
TF at the time TO of the time schedule of the transmission, which in fact corresponds to no object presented, the counter CMT will be effectively reset. However, if the user no longer presses the TF key until receipt of the JJ1 sound signal, this actuation will not be considered as the expression by the user of an interest in the object OB1, since the interval of time separating this premature actuation from the instant of reception of the sound signal Je 1 will be greater than the duration of the range PLT1 defined by the auxiliary indication contained in the sound signal JJ1.

Similarly, if the user presses the TF key several times within the range PLT1, the counter CMT will certainly be reset to zero each time, but, in any case, the time interval separating the last press of the TF key on receipt of the JJ 1 sound signal will make it possible to consider that the user has expressed an interest in the object
OB 1 in the range PLT 1.

 This advantage of the invention is also found in an application of the flow type of advertising messages. In fact, advertising messages are generally grouped in television programs.

However, only some of the messages, reserved for only some of the advertisers, can be associated with an offer to acquire the product presented, or with the sending of documentation. In other words, only certain of the advertising messages may comprise, typically at the end of the message, specific sound signals containing at least the identifier of the advertising message as well as the auxiliary indication making it possible to identify the duration of the message. In fact, in practice, in this case, the reference information will also include an identifier of the television channel and an identifier of the advertiser.

 The operation of the housing according to the invention is analogous to that which has been described above. In this case, however, in addition to the identifier of the advertising message, the identifier of the advertiser and the identifier of the television channel will be stored, as well as generally also the value of the additional counter making it possible to determine the time of day. 'pressing the activation keys, for example for statistical purposes.

 If a press of the TF key on the keyboard is made during an advertising message which is not a message concerned by the invention, two cases may arise. Or it is the last advertisement. In this case, after a predefined period, for example 5 minutes, and without reception of any specific sound signal, the content of the memory MM3 is automatically set to the value 0.

 Either this support is followed by another support located during an advertising message according to the invention, we then find ourselves in the operating cases explained above.

 If, on the other hand, this press is spaced temporally from the reception of a specific subsequent sound signal of a value greater than a duration of an advertising message, it will then be considered that this press does not correspond to the expression of an interest by the 'user.

 Finally, it is advantageously provided, whatever the applications, that the box will automatically go into standby mode if no key is pressed or if no specific sound signal is received for a predetermined duration, for example 5 minutes .

 At the end of the transmission, the viewer picks up the CT handset from his telephone set and dials the telephone number of the SV server.

When the communication is established, the viewer actuates for example another key TF2 of the keyboard, which has the effect of restoring outside the box a restitution information composed here of the box identifier and the content of the memory MM2 .

 In practice, in the example described here, the microcontroller PR includes coding means produced in software and intended to restore outside of the housing the restitution information in coded form. These coding means are activated in response to the actuation of the corresponding key on the keyboard. These coding means allow coding of the restitution information in the form of an acoustic bar code, which in particular makes it possible to avoid disturbing the transmission of the data with parasitic frequencies.

 Concretely, the coding means are capable, for a bit of logic value 1 of the restitution information, of delivering a logic signal having a high state for a first predetermined duration and a low state for a second predetermined duration. On the other hand, for a bit of logic value 0, the coding means deliver a logic signal having the low state during all of the first and second durations. In combination, the BZ transducer will not deliver any acoustic signal in response to a logic signal in the low state, and, in response to a logic signal in the high state, will deliver a single frequency acoustic signal whose frequency is substantially in the center of the bandwidth of a telephone line.

A frequency of 2048 Hz will therefore preferably be used, which is moreover easily usable from a conventional 4-bit microcontroller, since this frequency is a multiple of two. The logic signals S1 and S2 corresponding to bit codings having logic values 1 and 0 respectively, are illustrated in FIG. 4. The total duration T of transmission of a bit is 15.625 ms, which corresponds to approximately three times the time constant of a carbon microphone fitted to a telephone set. With regard to the signal S1, the duration
TT1 during which the logic signal is high, is taken equal to 11.71875 ms, while the duration TT2 is equal to 3.90652 ms.

 In fact, only the mask of the signal S1 is represented for the purposes of simplification in FIG. 4, and we will nevertheless speak of a high state for the signal S1 for the duration TT1 although in reality, during this duration, the signal consists of a plurality of peaks at the frequency of 2048 Hz.

 This transmission time of 15.625 ms, corresponding to a transmission speed of 64 baud, was adopted here because of its ease of implementation. However, it is possible to choose an optimal transmission speed between 15 baud and 16 baud, which makes it possible to obtain a stable transmission signal taking into account the time constant of a carbon microphone.

 Data transmission takes place by sending successive frames. The reception at the local server is then of the synchronous type, that is to say that there is a synchronization carried out on the header of each transmitted frame. This synchronization allows in particular the adjustment of the reception levels by an automatic gain control device, the detection of the transmission speed, and the setting of the reception means. The adjustment of the reception level by an automatic gain control device allows good reception of the frames and compensates for the loss of levels due to the carbon microphone, or by means of amplification of a signal produced by an electronic microphone.

A transmission frame is illustrated in FIG. 5. The duration of the header of a frame is 575 ms and is broken down into a logic signal.
S3 in the high state for a TT3 duration of 400 ms maximum allowing automatic gain control, followed by a signal in the low state S4 for a TT4 duration of 50 ms allowing the setting of the server receiver, finally followed by d '' a synchronization byte S5 comprising an alternation of bits 1 and bits 0 for 125 ms. The signals S1 and / or S2 corresponding to the restitution information are then transmitted.

The homologous means allowing the decoding of the restitution information are incorporated in particular in the interface
INT2 of the SV server.

 After reception by the interface INT2 of the server SV, of the acoustic signal containing the information of restitution, and extraction of the latter from this signal, the management means of the server compare the identifiers of objects (or advertising messages) ID1, ID2 with those contained in the BD database. When a comparison is positive, the management means then match the identifier associated with the corresponding object with the box identifier.

 The viewer's order is then recorded and he can later receive his object, or a documentation.

 The transmission identifier can be used to minimize the risk of error, while the instants T10, T20 can be used for statistical purposes.

 We will now describe in more detail with particular reference to FIGS. 6 to 15, an embodiment of the invention using for the transmission of reference information, a specific acoustic signal of the acoustic "Jingle" type.

 As illustrated in FIG. 6, two elementary acoustic signals SE1 or SE2 or sound elements, having equal elementary durations T equal to 65.536 ms, are used here to encode the reference information.

 As can be seen in this FIG. 6, the elementary acoustic signal SE2 is reversed in phase (that is to say multiplied by -1) with respect to the elementary acoustic signal SE1. In other words, and more simply, it will be said that the elementary acoustic signal, or sound element, SE1 is "direct" while the elementary acoustic signal SE2, or sound element, is "inverted".

 In FIG. 6, the sound elements have been represented in the form of a substantially sinusoidal signal, to simplify the graphic representations. This being so, in general, the invention is not limited to a particular type of sound element and allows the use of any sound element having a temporal evolution and any frequency spectrum.

 However, in the context of interactive television or radio programs, digital messages are intended to be transmitted over the air. However, the FM band (frequency modulation) is limited to around 15 kHz. This is the reason why sound elements whose frequency spectra do not contain frequencies above 15 kHz will preferably be used.

 In addition, the bandwidth of a current ordinary television speaker hardly extends beyond 8 to 10 kHz and can be reduced even more depending on the quality of the speakers used.

 Furthermore, although the presence of relatively low frequencies makes it possible to modify the tone of the signal and to make it more pleasant to the ear, it is preferable, for the decoding of the information, to eliminate as far as possible the maximum of vocal frequencies which could result from spoken human words, either during the formation of the elements of sound or when the reception of the signal containing the messages, due for example to people being near the microphone of the BT box.

 For all these reasons, it was considered preferable here to use a direct sound element, and consequently an inverted sound element, having a frequency spectrum whose most frequencies is between 1 kHz and 4 kHz.

 We can thus use for example the direct sound elements SE1 illustrated in FIGS. 7 and 8.

In FIG. 7, the sound element SE1 represented temporally partially over 22 ms has a frequency spectrum extending mainly between 1 kHz and 4 kHz with peaks of the order of 1.5; 2 and 3 kHz. Such a sound element makes it possible to obtain a specific sound signal
JJ1 or JJ2 analogous to the ringing of a mechanical alarm clock.

 In FIG. 8, the direct sound element SE1, partially represented over a duration of 27 ms, has a substantially sinusoidal shape and a frequency spectrum extending mainly between 1 and 2.5 kHz.

 In practice, these two elements of sound were obtained by calculation using a microcomputer and an associated sound card. More specifically, the sound element SE1 in FIG. 8 was obtained in particular from a frequency ramp and makes it possible to develop a specific sound signal JJ1 or JJ2 analogous to a chirping of a bird.

 That said, although the sound elements are generally generated by software, the invention does not exclude selecting a selected time range of a preexisting initial sound signal, for example a piece of music, to form said sound element. direct or reverse.

 As illustrated in FIG. 9, a bit of logical value 1 (for example) is coded from two identical consecutive sound elements, that is to say either as illustrated on the left side of FIG. 9, from two direct sound elements SE1, as illustrated on the right-hand side of FIG. 9 from the succession of two inverted sound elements SE2.

 As regards the coding of a bit of logical value 0, as shown in FIG. 10, the succession of two different sound elements is used. More precisely, one can use (left part of FIG. 10) a direct sound element SE1 followed by an inverted sound element in phase SE2 or else (right part of FIG. 10) the succession of an inverted sound element SE2 and a direct sound element SEl.

 While it would have been possible to use as many pairs of sound elements as bits to be transmitted, it was considered preferable, for reasons of compactness and therefore of decoding speed, to use, as illustrated on FIG. 11, N + 1 consecutive sound elements for coding N consecutive bits of a message. More precisely, in FIG. 11, the sequence 0 1 0 is coded from the four consecutive sound elements SE1, SE2, SE2 and SE1.

 As illustrated in FIG. 12, it has been considered preferable, in particular to facilitate the decoding of the information, that any message transmitted MS comprises an ETS synchronization header followed by the useful part PU proper of the message containing the information of reference and formed here of a byte.

 As will be explained in more detail below, it is particularly advantageous for the synchronization header to include at least 3 successive bits of different logical values, so as to be able to detect two logical transitions 0/1 and 1/0.

 In FIG. 13, it can be seen that the transmission device according to the invention, intended for transmitting acoustically via the radio link, an MS message, firstly comprises an MCO block incorporating means for generating the two elementary acoustic signals from which the different bits of the message will be coded, and coding means making it possible to assemble these different elements as a function of the logical values of the bits of the message. In a particular embodiment, one can use in this respect a microcomputer allowing the generation by calculation of the acoustic signal used to support the coding of binary information, and an associated sound card of the type well known to those skilled in the art. English name "Sound Blaster".

 In addition, there may be advantageously provided MXA analog mixing means perfectly known to those skilled in the art, and intended to incorporate into the specific sound signal containing the message (reference information) an additional SSA sound signal, for example music by low level background.

 The specific sound signal JJ 1 or JJ2 then containing the message MS to be transmitted, can be directly delivered to radio transmission means EMT of the station STM so as to be transformed into a radio signal intended to be broadcast via the antenna Au 1.

 It is also possible to record the signal JJ1 beforehand on an MTS recording means, for example an audio cassette, using a recorder / player device such as a tape recorder.

 However, in this case, the recording speed should be about the same as the replay speed except for one fluctuation. Indeed, a slowing down or an acceleration of the band results in a compression or a temporal dilation of the signal which can be troublesome for the decoding of the information such as it will be explained in more detail below. In other words, a tape recorder whose speed is not stable (which therefore gives rise to wailing and flickering), or whose frame rate is stable but too slow or too fast, can cause decoding.

 Generally, a means of recording and re-reading the specific sound signal will be chosen, the difference between the recording and reading speeds of which remains below a predetermined threshold which is linked to the maximum frequency contained in the signals. predetermined elementary elements (sound elements) as well as the elementary duration T of these sound elements. Thus, for example, for an elementary duration of a sound element equal to 65.536 ms, and for a sound element whose frequency spectrum is limited to 5 kHz, we will accept a maximum fluctuation in speed of the 0.04%.

 Likewise, for an elementary signal whose maximum frequency will not exceed 10 kHz, and always for an elementary duration of 65.536 ms, we will tolerate a maximum fluctuation between the playback and recording speeds, of the order of 0.02 %.

A person skilled in the art will therefore be able to choose a digital type recorder (DAT format: Digital Audio Tape) such as a professional tape recorder, for example the one sold by the
SONY company under the reference TCD - D7.

 The MRS means of the BT box comprise means for processing the received signal comprising, in the embodiment described here an analog stage MFA intended, as will be seen in more detail below, to perform a filtering of the received signal and a shaping of this last so as to deliver a binary SCB signal.

 The processing means also include autocorrelation means and message decoding means, distributed here between a specific integrated circuit (ASIC) referenced ASC and the processor PR.

 Generally, the decoding of the logical value of any bit of the message comprises, according to the invention, an autocorrelation of the signal received over an autocorrelation duration at most equal to the elementary duration of a sound element. In other words, we will compare, by this autocorrelation over a duration advantageously equal to the elementary duration, two adjacent time ranges of the received signal so as to determine whether these two ranges comprise two identical sound elements or else two sound elements opposite, which will make it possible to deduce the logical value of the bit thus received.

As illustrated more particularly in FIG. 14, the floor
MFA comprises two filters F1 and F2 which are here two bandpass filters centered around a chosen cutoff frequency, for example of the order of 3.4 kHz. These two bandpass filters are followed by a third filter
F3 which is in this case a high-pass filter having a cut-off frequency of the order of a few hundred Hz, typically 340 Hz.

 Indeed, the LV box is not connected by any link, in particular wired, to the television receiver, and this in particular for the purpose of simplicity of construction and cost reduction. Consequently, the microphone MC of the box BT receives not only the specific sound signal JJ1 (JJ2) containing the digital message, but also a certain number of ambient noises BAM, such as for example a bark of a dog or even conversation noises.

 Bandpass filtering therefore essentially has the function of attenuating ambient noise, and in particular any words spoken by people. Although it would have been possible to use only one bandpass filter, it has been considered preferable at present to use two filters so as to obtain a better stiffness of the resulting filtering and therefore a better attenuation of the undesirable frequencies, in particular the low vocal frequencies but also high frequencies which could generate an annoying whistling sound.

 Likewise, the high pass filter is not really essential.

This being so, it also contributes to eliminating the low frequencies and ensures in particular a suppression of the DC component of the signal.

 Following the filter F3, there is a comparator CMP1 intended to compare the filtered signal coming from the filter F3 with a reference value, in this case the zero value, so as to obtain a binary comparison signal, that is to say ie a signal capable of taking only two values, in this case the values +1 and -1. This binary comparison signal SCB is then delivered to the ASC ASIC (FIG. 15) to be sampled there by a conventional sampler ECH controlled by a clock signal CLK originating for example from a quartz QT. The sampling frequency chosen here is 3.9 kHz. Consequently, the SCB signal leaving the comparator will be sampled every 256 KILS.

 The samples corresponding to a duration equal to two elementary durations are stored in two shift registers R1 and R2 connected in cascade, each capable of storing 255 binary values. These registers therefore form a delay line.

 In fact, we have chosen here for the purpose of simplification of the registers each containing 255 memory boxes because each 256th pulse is used to effect the transfer of a sample in the first register is consequently a transfer of the last sample of the first register in the second register.

 Between two samples at 256 pLS, a bit-by-bit comparison between the two registers is carried out every microseconds by means of a set of OUEXCLUSIF gates referenced XOR.

 The autocorrelation of the received signal is therefore carried out here over a duration substantially equal to the elementary duration of 65.536 ms, to the nearest sample.

The output of this set of XOR doors is connected to the two inputs of a CPT counter. More precisely, for the index i, the value of the counter CPT is incremented if the two samples Rl (i) and
R2 (i) are equal and this counter is decremented if they are opposite.

 After comparing the 255 values, the state of the counter therefore reflects the resemblance between two adjacent time ranges of the received signal.

 Of course, a theoretical value of 255 would correspond to signals in perfect correlation and would therefore be representative of a bit 1 "of the message, while a value of -255 would correspond to signals in perfect opposition which would be equivalent to a bit" 0 "of the message.

 In practice, this counter value varies continuously between the two extreme values mentioned above.

 More precisely, a threshold S1 is then fixed for the decision to decode a bit "1" received and a threshold S2 is fixed for the decision to decode a bit "0" received. In practice, we have chosen these two thresholds respectively equal to +32 and -32.

 In other words, the counter output is compared to these different thresholds in a CMP2 hysteresis comparator, for example a Schmitt trigger, and the comparator output is connected to an BS output flip-flop.

In other words, after each comparison step the value of the counter is compared with the decision thresholds. The result of this operation is transmitted to the output flip-flop. If the counter value
CMP2 exceeds 32, the output of the BS flip-flop goes to 1. If the counter value becomes less than -32, the output of the BS flip-flop goes to 0. For all intermediate values, the output of the flip-flop remains unchanged.

 In fact, the decoding device according to the invention therefore behaves as if a window of width equal to two elementary durations were dragged over the signal to be analyzed. The BS output flip-flop is updated each time the decision threshold is crossed. If consecutive bits keep the same value, the SCF output of the output flip-flop does not change state.

 The SCF signal delivered by the output flip-flop is analyzed by the microprocessor PR.

 When a message is received, the processor PR will first of all detect the synchronization header of this message. In this regard, the software contained in the microprocessor will, throughout this synchronization period, continuously read the output of the flip-flop BS and detect any variation in value of this output, that is to say any logic transition 0/1 or 1/0.

 In addition to this logical transition detection, the processor will also analyze the duration separating two detected logical transitions and compare it with the elementary duration T of a sound element.

 Thus, when the processor has detected two successive logical transitions, for example the 0/1 transition then the 1/0 transition corresponding to the three bits of the synchronization header described above, and the duration separating these three bits thus detected will be approximately equal to 65.536 ms, a synchronization pulse will then be emitted signifying that the synchronization header has been recognized.

In the rest of the processing, it will no longer be necessary to read the output of the flip-flop BS continuously, but to perform this reading at particular instants spaced from the elementary duration, that is to say 65.536 ms, so as to read the autocorrelation value (bit state) corresponding to the middle of a sound element.

 During the synchronization phase, the comparison of the duration separating two detected logical transitions with the elementary duration makes it possible to differentiate the real bits of the bit synchronization header which would be detected and which would in fact be due only to parasites caused for example by ambient noise.

 In general, it has been observed that a sampling frequency at least equal to the highest frequency peak fmax remains acceptable in the application which has just been described. Indeed, although a sampling frequency fmax leads to loss of information for certain frequencies of the sound element in particular the frequency fmax, it turned out that this was not a problem for obtaining good decoding of the message taking into account that the selected sound element also has many lower frequencies than the sampling frequency fmax. Furthermore, the choice of such a sampling frequency has the advantage of limiting the complexity of the decoding circuits.

 This being so, the person skilled in the art can easily increase the sampling frequency of the SCB signal as a function of the nature of the sound element used and in particular if he uses sound elements exhibiting regular temporal evolutions such as for example sinusoids or damped sinusoids with well-defined unique frequencies.

Claims (12)

 1. Interactive method for transmitting information between a receiver reproducing a radio broadcast, a box having at least one manually operable control element, and a validation station, in which at least one group of is incorporated into the radio broadcast. specific information intended to be restored within a predetermined time range (PLT 1) and comprising a sequence of main information (IM 1) corresponding to a specific event occurring during the transmission, as well as reference information (JJ1) containing at least one identifier of the event and an auxiliary indication making it possible to define said time range, we store in the validation station (SV) validation information containing at least each event identifier, we receive each reference information (JJ1 ) at the level of the box (BT) and a storage indication con stored in the box holding at least the identifier of each specific event which gave rise to a detection within the housing of at least one actuation of the control element (TF) in the time range (PLT1) defined by the auxiliary indication associated with this event, restitution outside the box and to the validation station, restitution information containing an identifier (1DB) of the box and said storage indication, and analysis is carried out within the validation station ( SV), the feedback information received and the validation information stored, so as to possibly associate at least one event identifier with the box identifier.  2. Method according to claim 1, characterized in that each reference information (JJ1) is restored at the end of the corresponding time range (PLT1).  3. Method according to claim 2, characterized in that the detection of an actuation of the control element (TF) of the housing within a time range comprises the initialization of a time counter (CMT ) to a predetermined initial value on each actuation and the analysis, upon reception of the corresponding auxiliary indication, of the current value of the counter with respect to said time range taking into account said auxiliary indication.  4. Method according to one of the preceding claims, characterized in that the receiver (RCP) restores each reference information item within a specific sound signal (JJ1), and by the fact that the specific sound signal is received at the level of the box (BT), and said reference information is extracted therefrom.  5. Method according to one of the preceding claims, characterized in that the restitution information is restored to the outside of the housing in the form of an acoustic signal.  6. Method according to claim 5, in which the reference information is digital information made up of several bits, characterized in that a first elementary acoustic signal (SE 1) having predetermined characteristics is produced, and a second signal acoustic elementary (SE2) reversed in phase with respect to the first acoustic signal, the two elementary signals having equal elementary durations, and any bit having a first logical value is coded by a pair of first consecutive elementary acoustic signals or by a pair of second consecutive elementary acoustic signals, while each bit having a second logical value is coded by a pair of different consecutive elementary acoustic signals, and the specific sound signal is produced from the first and second elementary acoustic signals defining said reference information and by incorporated nt possibly additionally an additional sound signal (S SA), and by the fact that this specific sound signal (JJ1) is intended to be broadcast then broadcast by the receiver in said transmission.  7. Box, usable for interactive transmission of information between a receiver reproducing a radio broadcast, the box and a validation station, comprising means for receiving (MC, MRS) reference information capable of being reproduced by the receiver and containing at least one identifier of a specific event broadcast during the transmission and an auxiliary indication making it possible to define a broadcasting time range associated with this event, a first memory (mi1) containing a box identifier, an element of control (TF) capable of being actuated manually by a user of the housing, detection means (PR) capable of detecting any actuation of the control element, processing means (PR) capable, in the presence of at least an actuation detected in said time range, to store in a second memory (MM2) a storage indication containing at least the identifier of the specific event contained in said reference information, and controllable restitution means (PR, BZ) for restoring, outside the housing, restitution information containing the housing identifier and the storage indication.  8. Housing according to claim 7, characterized in that the detection means comprise a time counter (CMT) initializable to a predetermined initial value on each actuation of the control element (TF), and control means (PR ) able to read the current counter value (CMT) when receiving reference information.  9. Housing according to claim 8, characterized in that the detection means comprise a third memory (MM3) as well as control means (PR) able to store in the third memory (MM3) said actuation indication at each actuation of the control element.  10. Housing according to one of claims 7 to 9, characterized in that the reception means of the housing comprise a reception interface (MC) of a specific sound signal containing said reference information, and decoding means ( MRS, PR) capable of extracting said reference information from the sound signal.  11. Box according to one of claims 7 to 10, in which the reference information is digital, any bit of this reference information having a first logic value being coded by a pair of first consecutive predetermined elementary acoustic signals (SE1) or by a pair of second consecutive elementary acoustic signals (SE2), the second predetermined acoustic signal being reversed in phase with respect to the first elementary acoustic signal, while any bit having a second logic value is coded from a pair of signals different consecutive elementary acoustics (SE1, SE2), characterized in that the reception means comprise means (MFA) for shaping the specific sound signal received (JJ 1) comprising means for comparing (CMP 1) this signal to a reference value so as to obtain a binary comparison signal (SCB), sampling means (EC H) of the binary comparison signal, two registers (R1, R2) for storing the two groups of samples, and by the fact that the autocorrelation means comprise means of comparison of the EXCLUSIVE OR type (XOR) capable of performing a sample-to-sample comparison of the two groups of samples, a counter capable of accounting for the number of samples of identical or different values, and an output register (BS) capable of taking the logical value 0 or 1 as a function of the comparison of count value in relation to two predetermined thresholds.  12. Housing according to one of claims 7 to 11, characterized in that, said restitution information being made up of a plurality of bits, the restitution means comprise  - an activation element (TF2),  - coding means (PR) able, in response to actuation of the activation element, to deliver, for a bit of said restitution information having the logic value "one", a logic signal having a high state during a first predetermined duration and a low state for a second predetermined duration, and for a bit of said restitution information having the logical value "zero", a logic signal having the low state during all of the first and second durations, and  - an electroacoustic transducer (BZ) delivering no acoustic signal in response to a logic signal in the low state, and delivering in response to a logic signal in the high state, a monofrequency acoustic signal whose frequency is located substantially in the center of the bandwidth of a telephone line.