Monitoring system and method, corresponding computer program

Abstract

This acoustic monitoring system comprises a sensor (22) located in the area to be monitored and providing a first acoustic signal (S1), a local processor (24) for generating a second signal (S2) by configurable processing of the first signal, a local transmitter (26A) for transmitting the second signal outside the area to be monitored, and a remote monitoring server (20), located outside the area to be monitored, equipped with means (46) for analyzing the second signal to detect an anomaly. The remote monitoring server is further equipped with means (50) for generating a request (R) to modify at least one processing parameter of the first signal based on a result provided by its analysis means. A receiver (26B) coupled to the local processor is capable of receiving this request and transmitting it to the local processor for application of the required modification and transmission of the resulting second signal. Figure for the abbreviation: Fig. 2

Description

Title of the invention: Monitoring system and method, corresponding computer program

[0001] The present invention relates to a monitoring system, a monitoring method implemented by this system, and a computer program comprising instructions for executing such a monitoring method when executed by this system. It relates, for example, more specifically, but not exclusively, to the monitoring of activities of daily living (ADLs).

[0002] This type of monitoring system includes at least one sensor located in at least one area to be monitored, for example, the living space of at least one vulnerable and / or elderly person requiring close monitoring, in particular through the creation and maintenance of an activity log, and intervention in the event of the detection of a problem, anomaly, deterioration, or any significant change in circumstances affecting the safety of the person(s) concerned. Generally, the in-depth analysis of the signals captured for maintaining the activity log, or for detecting an anomaly and making a possible decision to intervene, is not performed locally but remotely by a monitoring server equipped with substantial computing resources, possibly shared between several areas to be monitored.The sensor is then integrated into a communicating or connected object, for example compatible with IoT (Internet of Things) technology, according to which this object is locally equipped with a sensor, limited processing means and means of transmitting data from the sensor to a remote server via the Internet network.

[0003] In practice, the sensor can be a simple acoustic sensor, a video surveillance camera, an infrared camera, a radar, or any other type of sensor capable of providing signals that can be used for surveillance as described above. In particular, for AVQ surveillance, an acoustic sensor can be advantageously used due to its simplicity and effectiveness, as sounds often carry highly relevant information about daily life activities without being overly intrusive on the privacy of the individuals being monitored, thus compromising their own safety.

[0004] The invention thus applies more particularly to a monitoring system comprising: - at least one sensor placed in at least one area to be monitored and providing at least one initial signal carrying information; - at least one local computer coupled to said at least one sensor, configurable using at least one processing parameter of said at least one first signal for the generation of at least one second signal to be transmitted, carrying at least part of the information; - at least one transmitter coupled to said at least one local computer, capable of transmitting said at least one second signal outside said at least one area to be monitored; and - a remote monitoring server, located outside said at least one area to be monitored, equipped with means of receiving said at least one second signal and means of analyzing said at least one second signal for the detection of at least one anomaly or for keeping an activity log.

[0005] The main problem posed by such a monitoring system, with at least one local communicating object (i.e., the sensor itself, the local computer, and the transmitter coupled to the local computer) and remote analysis (i.e., the remote monitoring system), is finding a compromise between the volume of information to be transmitted and the computing power to be used locally. Indeed, if no local computing power is required, then the entirety of at least one initial signal must be transmitted as is to the remote monitoring server, which is very expensive in terms of transmission. Alternatively, only a portion of the signal can be transmitted, but this risks indiscriminately missing a potentially important event.Conversely, if the aim is to minimize transmission costs, then local processing of at least one initial signal before transmission is necessary to reduce its volume as much as possible. This results in a significant loss of information and degraded analysis by the remote server, which must be compensated for by a necessarily costly local analysis capacity.

[0006] Today, one initial solution consists of "blindly" reducing the volume of acoustic information to be transmitted. The local computer(s) is / are then equipped with known and more or less complex means of selecting the information to be transmitted, for example, according to an absolute or relative signal level threshold, time sampling, periodic selection, frequency filtering, digital quantization, compression, etc. This at least one second signal thus obtained is certainly potentially of significantly lower volume than the at least one first signal provided by the sensor(s), but the relevance of the transmitted information is neither truly known nor controlled. Moreover, this sometimes requires computing resources that the at least one local communicating object does not possess.

[0007] A second solution consists of varying the resolution (i.e., by time sampling, frequency filtering, digital quantization, compression) according to from a local analysis of the information in said at least one first signal. This is, for example, the teaching of patent document WO 2020 / 092701 A2, according to which said at least one second signal is transmitted in high or low resolution depending on whether a predetermined event is detected or not in a portion of the information, in particular that which remains in said at least one second signal in low resolution. Not only is the relevance of the switching criterion between low and high resolution not guaranteed, but it also requires even greater computing power, which said at least one local communicating object does not always possess.

[0008] It may therefore be desirable to provide an acoustic monitoring system which makes it possible to overcome at least some of the aforementioned problems and constraints.

[0009] A monitoring system is therefore proposed comprising: - at least one sensor placed in at least one area to be monitored and providing at least one initial signal carrying information; - at least one local computer coupled to said at least one sensor, configurable using at least one processing parameter of said at least one first signal for the generation of at least one second signal to be transmitted, carrying at least part of the information; - at least one transmitter coupled to said at least one local computer, capable of transmitting said at least one second signal outside said at least one area to be monitored; and - a remote monitoring server, located outside of said at least one area to be monitored, equipped with means for receiving said at least one second signal and means for analyzing said at least one second signal for the detection of at least one anomaly or for keeping an activity log; in which: - the remote monitoring server is also equipped with means for generating at least one request to modify said at least one processing parameter of said at least one first signal based on a result provided by its analysis means; - the monitoring system further comprises at least one receiver coupled to said at least one local computer, capable of receiving said at least one modification request and transmitting it to said at least one local computer for application of the required modification and transmission of said at least one resulting second signal.

[0010] Thus, the monitoring system has a feedback loop to adjust the processing of the signal or signals to be transmitted according to the analysis needs of its remote monitoring server, whose analysis capabilities are logically more powerful than those of said at least one local computer. This results in a better relevance of the processing performed on said at least one first signal since this processing is then defined by and according to the expectations of the remote server.

[0011] Optionally: - said at least one local computer includes at least one buffer memory for recording and storing said at least one first signal for at least a predetermined sliding time window; and - the means of generating said at least one modification request are capable of generating a request to send at least part of the information of said at least one first signal stored in said at least one buffer memory in said at least one predetermined sliding time window according to a modified processing parameter of said at least one local computer.

[0012] Optionally also, the means for generating said at least one modification request are further capable of generating a modification request for recording and storing parameters of at least a part of said at least one first signal in said at least one buffer.

[0013] Optionally, said at least one request to modify recording and storage parameters of said at least one first signal in said at least one buffer includes at least one of the elements of the set consisting of: - a modification of said at least a predetermined sliding time window; - a modification of the temporal sampling frequency to be carried out on said at least one first signal for its storage in said at least one buffer memory; - a change in a criterion triggering the recording and storage of said at least one first signal in said at least one buffer memory; - a frequency filtering modification of said at least a first signal, defining in particular a range of acoustic frequencies, for its storage in said at least a buffer memory; - a modification of the digital quantization of said at least one first signal for its storage in said at least one buffer memory; - a compression modification of said at least one first signal for its storage in said at least one buffer memory.

[0014] Optionally, said at least one processing parameter includes at least one of the elements of the set consisting of: - an absolute or relative signal level threshold, for example expressed in dB when it is a sound level, beyond which the acoustic information of the first signal can be transmitted by the second signal; - a similarity threshold, for example measured in a spectrogram of said at least one first signal, from which it is decided whether the information of said at least one first signal can or cannot be transmitted by said at least one second signal; - a temporal sampling frequency of said at least a first signal for obtaining said at least a second signal; - a periodic transmission parameter; - at least one frequency filtering parameter, defining in particular a range of acoustic frequencies to be transmitted, of said at least one first signal for obtaining said at least one second signal; - at least one digital quantization parameter of said at least one first signal for obtaining said at least one second signal; - at least one compression parameter of said at least one first signal for obtaining said at least one second signal; - at least one parameter for the duration of the acoustic information of said at least one first signal to be transmitted in said at least one second signal; and - at least one parameter for interrupting the transmission of said at least one second signal until a predetermined condition detectable by said at least one local computer is again met, or for resuming transmission of the second signal.

[0015] A monitoring method is also proposed, comprising the following steps: - reception, by at least one local computer coupled to at least one acoustic sensor located in at least one area to be monitored, of at least one first signal carrying acoustic information provided by said at least one acoustic sensor; - processing of said at least one first signal for the generation of at least one second signal to be transmitted, carrying at least part of the information, by said at least one local computer, configurable using at least one processing parameter of said at least one first signal; - transmission of said at least a second signal outside said at least one area to be monitored by at least one transmitter coupled to said at least one local computer; and - reception and analysis of said at least a second signal by a remote monitoring server, located outside said at least one area to be monitored, for the detection of at least one anomaly or for the maintenance of an activity log; including the following steps: - generation, by the remote monitoring server, of at least one request to modify said at least one processing parameter of said at least one first signal based on a result provided by its analysis of said at least one second signal; - receipt of said at least one modification request by at least one receiver coupled to said at least one local computer; - application of the modification required by said at least one local computer; and - transmission of said at least a second resulting signal by said at least a transmitter coupled to said at least a local computer.

[0016] Optionally: - said at least one second signal is transmitted outside said at least one area to be monitored with a timestamp of each piece of information it contains; and - said at least one request to modify said at least one processing parameter of said at least one first signal is contextualized in that it includes at least one duration preceding and / or following the timestamp of a particular information detected during the analysis of said at least one second signal for a return of said at least one second signal in this at least one duration and according to a higher quality.

[0017] Optionally, said at least one request to modify said at least one processing parameter of said at least one first signal includes: - a request to interrupt the transmission of said at least a second signal until a predetermined condition, detectable by said at least one local computer, is again met; or - a request to resume transmission of the second signal.

[0018] Optionally also, the predetermined condition detectable by said at least one local computer includes a predetermined duration and / or a detectable condition in said at least one first signal.

[0019] A computer program downloadable from a communication network and / or stored on a computer-readable medium and / or executable by a processor is also proposed, characterized in that it includes instructions for executing the steps of a monitoring method according to the invention, when said program is executed by at least one local computer coupled to at least one sensor acoustic and by a remote monitoring server of an acoustic monitoring system.

[0020] The invention will be better understood with the aid of the following description, given solely by way of example and made with reference to the accompanying drawings in which: - [Fig. 1] schematically represents the general structure of a surveillance system according to one embodiment of the invention, - [Fig.2] schematically and in more detail represents the functional structure of a communicating object with an acoustic sensor and a remote monitoring server of the monitoring system of [Fig.1], - [Fig.3] illustrates the successive steps of a monitoring process implemented by the system of [Fig.1].

[0021] The installation schematically represented in [Fig. 1] illustrates the main room of a monitored area 10, for example, an apartment for an elderly person 12 in a specialized group residence. This main room has two doors: one 14 for access to a bathroom that is part of the monitored area 10, and the other 16 for exiting the monitored area 10 and accessing the common areas of the group residence. This is only a non-limiting example of an application context for the present invention. The invention also applies to the monitoring of several areas and to the monitoring of several people in one or more areas.

[0022] This installation further illustrates a monitoring system comprising a communicating object 18 in telecommunication with a remote monitoring server 20. Several communicating objects could also be envisaged.

[0023] The communicating object 18 itself comprises a sensor, for example an acoustic sensor 22 such as a microphone, a local computer 24 coupled to the acoustic sensor 22, and a transmitter / receiver 26 coupled to the local computer 24. It is located in the area to be monitored 10, for example in the main room, particularly on the ceiling. As already mentioned previously, any other type of sensor (video, infrared, radar, etc.) could be considered, either instead of or in addition to the acoustic sensor 22.

[0024] The remote monitoring server 20 is located outside the monitored area 10. It can be shared by several communicating devices located in different apartments within the specialized residence and can be placed in a dedicated monitoring room. In this case, it is advantageously connected to the various communicating devices in the residence via a local telecommunications and / or data transmission network. It can also be shared by several communicating devices in several apartments across several residences. In this case, it is connected to the various communicating devices via a wide area telecommunications and / or data transmission network such as the Internet 28. In all cases, it is equipped with means to keep an activity log and / or trigger an alarm or alert a response team 30 when an anomaly is detected.

[0025] The monitoring system 18, 20 of [Fig.1] is functionally detailed in [Fig.2],

[0026] As illustrated in this figure, the acoustic sensor 22 of the communicating object 18 provides a first signal SI carrying acoustic information. This signal is optimal in terms of content, since it is directly provided by the acoustic sensor 22 without any post-processing. Its quality therefore depends only on the acoustic sensor 22 itself. It is provided at each instant t, according to a predetermined optimal temporal sampling or continuously if possible, and the value it takes at each instant t is denoted Sl(t).

[0027] The local computer 24 of the communicating object 18 is advantageously designed to consume little energy, particularly according to IoT technology. It therefore has limited computing power and memory compared to that of the remote monitoring server 20. Apart from this, it can have a completely conventional architecture with a processing unit 32 (for example a processor) associated with at least one memory 34 (for example RAM or other) for storing data files and computer programs whose instructions are intended to be executed by the processing unit 32.The local computer 24 thus functionally comprises several computer programs or several functions of the same computer program to perform configurable processing of the first signal S1 and provide, as output from this processing, a second signal S2 to be transmitted, carrying at least part of the acoustic information of the first signal SL. These computer programs or computer program functions are presented as multiple and distinct, but this distinction is purely functional. They can just as easily be grouped in all possible combinations into one or more software programs. Their functions could also be at least partially microprogrammed or micro-wired into dedicated integrated circuits.Thus, as an alternative, the computer device with processing unit 32 and memory 34 implementing the local computer 24 could be replaced by an electronic device composed solely of digital circuits (without a computer program) for the performance of the same functions.

[0028] Optionally but advantageously, the local computer 24 includes a buffer memory 36, for example included in the memory 34, for recording and storing at maximum resolution the first signal SI provided by the acoustic sensor, in whole (SI) or in part (SI'), during a predetermined sliding time window denoted W, for example 5 minutes. At each instant t, the signal SI or SI' stored in buffer 36 therefore extends from Sl(tW) to Sl(t), or from Sl'(tW) to Sl'(t).

[0029] The parameter for recording and storing the first SI signal in buffer memory 36 includes, for example, at least one of the following parameters: - a parameter defining the sliding time window W; - the frequency of a time sampling to be performed on the first SI signal for its storage in buffer memory 36, for example a frequency deemed optimal and maximum of 48 kHz; - a criterion triggering the recording and storage of the first SI signal in buffer memory 36, such as an absolute or relative threshold of minimum sound level, for example expressed in dB; - at least one frequency filtering parameter, defining in particular an extended range of acoustic frequencies of the first SI signal to be recorded and stored in buffer memory 36; - at least one digital quantization parameter of the first signal S1 for its storage in buffer memory 36; - at least one compression parameter of the first signal SI for its storage in buffer memory 36, for example according to a minimum compression of the standard MP3 or Ogg format, or according to a lossless compression of the FLAC type (from the English "Free Lossless Audio Coded").

[0030] In accordance with the general principles of the present invention, the transmitter / receiver 26 of the communicating object 18 can functionally be subdivided into a transmitter 26A, for the transmission of the second signal S2 outside the area to be monitored 10, and a receiver 26B, for the reception of a request R to modify at least one processing parameter of the first signal SI, this request coming from outside the area to be monitored 10, in particular from the remote monitoring server 20.

[0031] As further illustrated in [Fig.2] and also in accordance with the general principles of the present invention, the monitoring server 20 is equipped with a receiver 38A, for receiving the second signal S2 transmitted by the communicating object 18, and a transmitter 38B, for transmitting the modification request R to the communicating object 18.

[0032] It further comprises a remote computer 40, advantageously equipped with significantly greater computing power and memory than the local computer 24. Apart from this, the remote computer 40 may have a completely conventional architecture with a processing unit 42 (for example, a processor) associated with at least one memory 44 (for example, RAM or other memory). It may, for example, be implemented in a computing device such as a conventional computer comprising a processor associated with one or more memories for storing Data files and computer programs whose instructions are intended to be executed by the processing unit 42. As illustrated in [Fig. 2], the remote computer 40 thus functionally comprises three computer programs 46, 48, 50, or three functions of the same computer program. It should be noted that the computer programs 46, 48, 50 are presented as distinct, but this distinction is purely functional. They could just as easily be grouped in all possible combinations into one or more software programs. Their functions could also be at least partially microprogrammed or micro-wired into dedicated integrated circuits. Alternatively, the computer system implementing the remote computer 40 could be replaced by an electronic device composed solely of digital circuits (without a computer program) to perform the same functions.Alternatively, at least some of the aforementioned computer programs could be remote and accessible by the remote computer 40 via the Internet. Generally speaking, even if all the aforementioned software and memory components are presented as being contained within the same remote computer 40, they could just as easily be dispersed across separate hardware components, even located far from each other, but interconnected via a network (data transmission bus, local area network, wide area network, Internet, etc.).

[0033] The computer program 46 includes instructions which, when executed by the processing unit 42, perform an analysis of the second signal S2 for the detection of at least one predetermined anomaly or simply for maintaining an activity log. This analysis as such is not part of the subject matter of the present invention. It may, in particular, make use of an artificial intelligence software module providing assistance in anomaly detection and decision-making, notably to alert the intervention team 30. One of the results of this analysis may therefore be the detection, with a sufficiently high estimated probability, of an indication of one or more anomalies in the signal S2, or the extraction, with a sufficient estimated quality, of relevant information to populate the activity log. In this first case, following the execution of the computer program 46, the computer program 48 may advantageously be executed.Another result of this analysis may be the observation of insufficient acoustic information carried by the second signal S2, such that at least one anomaly detection cannot be validated with sufficient probability, or that an intervention decision cannot be made, or that the extracted information is not of sufficient quality to populate the activity log. In this second case, following the execution of computer program 46, computer program 50 is executed.

[0034] The computer program 48 includes instructions which, when executed by the processing unit 42, trigger an alert (audible, visual, via messaging, ...) and / or a request for action from the intervention team 30 depending on the detected anomaly and / or a simple insertion of follow-up information into the activity log. Regarding the input of the activity log, the computer program 48 can be designed to detect classes of sounds, for example, those related to specific activities such as footsteps, trips to the toilet, coughing, breathing patterns during sleep, etc. Its primary function is to collect long-term activity data that can aid in monitoring and ensuring the safety of the individual concerned (i.e., detection or monitoring of a pathology, assessment of a level of autonomy, detection of a decline in health, etc.).This activity log is also intended for doctors, family members, or caregivers, the latter of whom may be interested although they are more often called upon for a one-off intervention or to visit outside of sleeping hours, for example.

[0035] The computer program 50 includes instructions which, when executed by the processing unit 42, generate the request R to modify at least one processing parameter of the first signal SI to obtain the second signal S2, depending on the insufficiency result provided by the execution of the computer program 46.

[0036] This processing parameter to be modified can be one or more of the following parameters: - an absolute or relative sound level threshold, for example expressed in dB, beyond which the acoustic information of the first signal SI can be transmitted by the second signal S2; - a similarity threshold, for example measured in a spectrogram of the first signal SI, from which it is decided whether the information of the first signal S1 can or cannot be transmitted by the second signal S2: it is thus verified whether the first signal SI is at each instant sufficiently different from its previous values ​​to trigger a transmission of the signal S2, the similarity measurement being done for example by studying the patterns present in the spectrogram of the signal SI; - a time sampling frequency of the first signal SI for obtaining the second signal S2, for example between 4 and 48 kHz, with a default frequency of 16 kHz considered reasonable; - a periodic transmission parameter; - at least one frequency filtering parameter, defining in particular a range of acoustic frequencies to be transmitted, of the first signal S1 for obtaining the second signal S2; - at least one digital quantization parameter of the first signal S1 for obtaining the second signal S2; - at least one compression parameter for the first signal SI to obtain the second signal S2, for example according to any of the possible compression ratios of the standard MP3 or Ogg format; and - at least one parameter for the duration of the acoustic information of the first signal S1 to be transmitted in the second signal S2.

[0037] It is observed that at least part of the processing parameters of the first signal SI for obtaining the second signal S2 can be similar to the aforementioned recording and storage parameters.

[0038] As regards the choice of modification of this or that processing parameter according to the insufficiency analyzed in the S2 signal, this is very dependent on the application context and within the reach of the person skilled in the art.

[0039] Thus, if a quality deficiency is detected in the second signal S2 by execution of computer program 46 or if a supplement is deemed necessary, the execution of computer program 50 may result in a request to reduce the absolute or relative sound level threshold from which the acoustic information of the first signal S1 must be transmitted by the second signal S2, and / or to increase the temporal sampling frequency of the first signal S1, and / or to modify or even eliminate the periodic transmission parameter (for example, the requirement for at least a certain duration of a transmission at every instant of the second signal S2 without a particular periodicity), and / or to increase the range of acoustic frequencies to be transmitted, and / or to reduce the compression of the first signal S1,and / or to a revision of the duration of the acoustic information of the first signal S1 to be transmitted or retransmitted in the second signal S2.

[0040] In particular, thanks to the portion of signal S1 recorded and stored at maximum resolution in buffer memory 36, at least a part of this maximum-resolution signal within the sliding time window W, [Sl(tW) ; Sl(t)] or [Sl'(tW) to S1'(t)], can be sent back with the application of the modified processing parameter(s) described above in the desired time interval and duration. In this case, the local computer 24 retrieves the desired portion from the signal portion S1 stored in buffer memory 36 and applies to it the modified processing parameter(s) in the request R received by receiver 26B to obtain and send or resend the resulting second signal S2.

[0041] Optionally but advantageously, the R query may include the modification of at least one of the aforementioned parameters for recording and storing at least a part SI' of the first signal SI in buffer 36, whether for the purpose of increase the resolution, up to possibly reaching the storage of the entire first signal SI in the defined sliding time window W, or to reduce it, for example in order to increase this sliding time window W given the limited capacities of the buffer memory 36.

[0042] Optionally, but advantageously, the second signal S2 is transmitted with a timestamp of each acoustic information it contains. The request R can then be contextualized in that it includes at least one duration preceding and / or following the timestamp of a particular acoustic information detected during the analysis of the second signal S2 by the remote monitoring server 20, for the retransmission of the second signal S2 within this at least one duration and with a higher quality to be defined by the desired modification of the aforementioned processing parameters.

[0043] Optionally, but advantageously, the R request can also follow a logic opposite to that presented previously by reducing the quality of the S2 signal to be transmitted. For example, if the analysis performed by the remote monitoring server 20 detects no anomalies for a certain time or no relevant monitoring information, it can include a request to interrupt the transmission of the second signal S2 until a predetermined condition, easily detectable by the local computer 24 of the communicating object 18, is met again. This predetermined condition can include a predetermined duration and / or a condition detectable in the first signal S1.

[0044] The operation of the monitoring system 18, 20 of figures 1 and 2 will now be detailed with reference to [Fig.3].

[0045] During a first reception step 100 which is executed at each instant t, the local computer 24 of the communicating object 18 receives the first signal SI provided by the acoustic sensor 22.

[0046] During a temporary backup step 102, executed when the local computer 24 has buffer memory 36 available and also at any given time, the signal SI is recorded and stored in buffer memory 36 in full SI form or slightly degraded SI' but at maximum resolution, applying the aforementioned recording and storage parameters. If buffer memory 36 is full, storage is performed by overwriting the oldest data, i.e., according to a FIFO (First In First Out) data deletion logic.

[0047] During a processing step 104, executed at each instant and in parallel with step 102, the first signal SI is processed by applying the aforementioned processing parameters to obtain the second signal S2 to be transmitted.

[0048] During a subsequent transmission step 106, also executed at each instant and / or under the conditions defined by the periodic transmission parameter or that of the aforementioned duration, the second signal S2 is transmitted by the transmitter 26A of the communicating object 18 to the receiver 38A of the remote monitoring server 20.

[0049] During a subsequent reception and analysis step 108, executed at each instant, the second signal S2 is received by the receiver 38A and analyzed by execution of the computer program 46.

[0050] During a subsequent alert step 110, if the execution of computer program 46 leads to the detection of at least one anomaly, then the execution of computer program 48 is triggered to issue an alarm or a request for intervention from team 30. If it simply leads to the extraction of follow-up information deemed sufficiently relevant, then the execution of computer program 48 is also triggered to insert this information into the activity log.

[0051] During a correction step 112 following step 108 and which can be executed independently of step 110, if the execution of the computer program 46 leads to the detection of at least one deficiency in the second signal S2, for example preventing the detection of an anomaly or absence of anomaly with certainty, or preventing the extraction of relevant tracking information, then the execution of the computer program 50 is triggered for the generation of the request R to modify the aforementioned parameters of processing the first signal S1 in order to obtain the second signal S2, this request R being able to optionally include in addition a request to modify the parameters of recording and storing the first signal S1 in the buffer memory 36.

[0052] During a subsequent transmission step 114, the request R is transmitted by the transmitter 38B and received by the receiver 26B of the communicating object 18.

[0053] During a subsequent modification step 116, the request R is transmitted by the receiver 26B to the local computer 24 to which it is coupled for modification of the processing of the first signal SI in preparation for the transmission or retransmission of the resulting second signal S2. This step is therefore followed by a return to step 104.

[0054] According to this operation, in a first example of a possible scenario, a sound characteristic of a worrying situation, which could correspond to the fall of person 12 or a heavy object, is detected by the remote computer 40 in the second signal S2. To confirm or refute this worrying situation, the remote computer 40 generates and sends a request R to enhance the quality of the second signal S2 to be received from the detected characteristic sound, according to at least one of the aforementioned processing parameters, so as to be able to detect whether this sound is followed, for example, by a whimper, shortness of breath, a groan (confirmation of the worrying situation) or whether, on the contrary, it is followed, for example, by footsteps, a predetermined sound ("darn" or equivalent), or any other action aimed at to reassure about the consequences of the detected characteristic sound. In addition, the R request can request the retransmission of the S2 signal in better resolution in the recent past of the detection of the characteristic sound, for example the last ten seconds if this is possible given the capacities of buffer memory 36, for a better analysis of the moments that preceded the detection of this sound.

[0055] According to another possible scenario, the breathing patterns of person 12 during sleep are analyzed at least periodically by the remote computer 40 in the second signal S2, to potentially detect a sleep apnea problem. It can then, at any time, generate and send a request R to enhance the quality of the second signal S2 to be received for a few minutes, in order to obtain a continuous and detailed recording allowing it to further analyze the data if necessary.

[0056] According to another possible scenario, for example in a case where the transmission of the second signal S2 to the remote monitoring server 20 is based on a parameterized sound level sufficient for monitoring the exit of a monitored area 10 of person 12, the footsteps of person 12 or the opening / closing of one of the doors 14, 16 trigger(s) the transmission of the second signal S2. After analysis performed by executing the computer program 46, a modification request R can be generated and issued for sending / re-sending the acoustic information following and preceding the time-stamped triggering event, even if it is lower than the initially parameterized sound level.Analyzing this acoustic information then allows us to better understand the situation by subsequently detecting, for example, footsteps moving away (indicating an exit from the area to be monitored through door 16) and thus alerting the intervention team 30, or on the contrary, a flushing noise indicating access through door 14 to the sanitary room which is part of the area to be monitored 10.

[0057] Many other scenarios can be imagined based on the operation described above.

[0058] It is clear that an acoustic monitoring system such as the one described above makes it possible to optimize the transmission and remote analysis of acoustic information provided by an acoustic sensor with limited local computing capabilities.

[0059] The feedback loop constituted by the possibility of sending back requests to modify the processing parameters of the first signal SI provided by the acoustic sensor to obtain the second signal S2 to be transmitted to the remote monitoring server allows the latter to regulate flexibly and on demand the information it really needs at any given moment.

[0060] Furthermore, the presence of a buffer alongside the acoustic sensor for temporary storage of the first signal it provides allows the feedback loop to work backward in time, within the limits of this buffer's capacity, to refine the analysis performed by the remote monitoring server. It also helps manage latency issues that might arise related to transfer and analysis times.

[0061] It should also be noted that the invention is not limited to the embodiment described above, the latter having been preferred for its simplicity of presentation.

[0062] Alternatively, several local communicating objects, with or without sensors of different types, can interact with the remote monitoring server. The latter can process several secondary signals in parallel and manage one or more modification requests for one or more local computers. Thus, for example, several sensors can be deployed in the same location, their secondary signals being centralized by the monitoring server to send a single modification request adapted to all the local computers: a corresponding application example is fire or intrusion detection, where a suspicious noise detected by one of the sensors justifies putting all the sensors on alert.

[0063] It will more generally be apparent to a person skilled in the art that various modifications can be made to the embodiment described above, in light of the instruction just disclosed to them. In the detailed presentation of the invention given previously, the terms used should not be interpreted as limiting the invention to the embodiment set forth in this description, but should be interpreted to include all equivalents that a person skilled in the art can foresee by applying their general knowledge to the implementation of the instruction just disclosed to them.

Claims

Demands

1. Surveillance system comprising: - at least one sensor (22) placed in at least one area to be monitored (10) and providing at least one first signal (SI) carrying information; - at least one local computer (24) coupled to said at least one sensor (22), configurable using at least one processing parameter of said at least one first signal (SI) for the generation of at least one second signal (S2) to be transmitted, carrying at least part of the information; - at least one transmitter (26) coupled to said at least one local computer (24), capable of transmitting said at least one second signal (S2) out of said at least one area to be monitored (10); and - a remote monitoring server (20), located outside said at least one area to be monitored (10), equipped with means (3 8A) for receiving said at least one second signal (S2) and means (46) for analyzing said at least one second signal (S2) for the detection of at least one anomaly or for keeping an activity log; characterized in that: - the remote monitoring server (20) is further equipped with means (50) for generating at least one request (R) to modify said at least one processing parameter of said at least one first signal (SI) based on a result provided by its analysis means (46); and - the monitoring system further comprises at least one receiver (26B) coupled to at least one local computer (24), capable of receiving at least one modification request (R) and transmitting it to at least one local computer (24) for application of the required modification and transmission of at least one resulting second signal (S2).

2. A monitoring system according to claim 1, wherein:

3.

4. - said at least one local computer (24) includes at least one buffer memory (36) for recording and storing said at least one first signal (SI) during at least one predetermined sliding time window; and - the means (50) for generating said at least one modification request (R) are capable of generating a request to send at least part of the information of said at least one first signal (SI) stored in said at least one buffer memory (36) in said at least one predetermined sliding time window according to a modified processing parameter of said at least one local computer (24). Monitoring system according to claim 2, wherein the means (50) for generating said at least one modification request (R) are further capable of generating a modification request for recording and storing parameters of at least a part of said at least one first signal (SI) in said at least one buffer memory (36). A monitoring system according to claim 3, wherein said at least one request (R) for modifying recording parameters and storing said at least one first signal (SI) in said at least one buffer memory (36) comprises at least one of the elements of the set consisting of: - a modification of said at least a predetermined sliding time window; - a modification of the temporal sampling frequency to be carried out on said at least one first signal (SI) for its storage in said at least one buffer memory (36); - a change in a criterion triggering the recording and storage of said at least one first signal (SI) in said at least one buffer memory (36); - a frequency filtering modification of said at least a first signal (SI), defining in particular a range of acoustic frequencies, for its storage in said at least a buffer memory (36);

5. - a modification of the digital quantization of said at least one first signal (SI) for its storage in said at least one buffer memory (36); - a compression modification of said at least one first signal (SI) for its storage in said at least one buffer memory (36). A monitoring system according to any one of claims 1 to 4, wherein said at least one processing parameter comprises at least one of the elements of the assembly consisting of: - an absolute or relative signal level threshold, for example expressed in dB when it is a sound level, beyond which the information of said at least one first signal (SI) can be transmitted by said at least one second signal (S2); - a similarity threshold, for example measured in a spectrogram of said at least one first signal (SI), from which it is decided whether the information of said at least one first signal (SI) can or cannot be transmitted by said at least one second signal (S2); - a temporal sampling frequency of said at least one first signal (S1) for obtaining said at least one second signal (S2); - a periodic transmission parameter; - at least one frequency filtering parameter, defining in particular a range of acoustic frequencies to be transmitted, of said at least one first signal (S1) for obtaining said at least one second signal (S2); - at least one digital quantization parameter of said at least one first signal (S1) for obtaining said at least one second signal (S2); - at least one compression parameter of said at least one first signal (S1) for obtaining said at least one second signal (S2); - at least one duration parameter for the information in said at least one first signal (S1) to be transmitted in said at least one second signal (S2); and

6. - at least one transmission interruption parameter of said at least one second signal (S2) until a predetermined condition detectable by said at least one local computer (24) is again verified, or resumption of transmission of the second signal (S2). Monitoring procedure, comprising the following steps: - reception (100), by at least one local computer (24) coupled to at least one sensor (22) located in at least one area to be monitored (10), of at least one first signal (SI) carrying information provided by said at least one acoustic sensor (22); - processing (104) of said at least one first signal (SI) for the generation of at least one second signal (S2) to be transmitted, carrying at least part of the information, by said at least one local computer (24), configurable using at least one processing parameter of said at least one first signal (SI); - transmission (106) of said at least one second signal (S2) out of said at least one area to be monitored (10) by at least one transmitter (26A) coupled to said at least one local computer (24); and - reception and analysis (108) of said at least a second signal (S2) by a remote monitoring server (20), located outside said at least one area to be monitored (10), for the detection of at least one anomaly or for the maintenance of an activity log; characterized in that it also includes the following steps: - generation (112), by the remote monitoring server (20), of at least one request (R) to modify said at least one processing parameter of said at least one first signal (S2) according to a result provided by its analysis (108) of said at least one second signal (S2); - receipt (114) of said at least one modification request (R) by at least one receiver (26B) coupled to said at least one local computer (24); - application (104) of the modification required by said at least one local computer (24); and - transmission (106) of said at least a second signal (S2) resulting from said at least one transmitter (26A) coupled to said at least one local computer (24).

7. A monitoring method according to claim 6, wherein: - said at least one second signal (S2) is transmitted out of said at least one area to be monitored (10) with a timestamp of each piece of information it contains; and - said at least one request (R) to modify said at least one processing parameter of said at least one first signal (SI) is contextualized in that it includes at least one duration preceding and / or following the timestamp of a particular piece of information detected during the analysis (108) of said at least one second signal (S2) for a return of said at least one second signal (S2) in this at least one duration and in a higher quality.

8. A monitoring method according to claim 6, wherein said at least one request (R) to modify said at least one processing parameter of said at least one first signal (SI) comprises: - a request to interrupt transmission of said at least one second signal (S2) until a predetermined condition detectable by said at least one local computer (24) is again verified; or - a request to resume transmission of the second signal (S2).

9. A monitoring method according to claim 8, wherein the predetermined condition detectable by said at least one local computer (24) comprises a predetermined duration and / or a detectable condition in said at least one first signal (SI).

10. A computer program downloadable from a communication network and / or stored on a computer-readable medium and / or executable by a processor, characterized in that it includes instructions for executing the steps of a monitoring method according to any one of claims 6 to 9, when said program is executed by at least one local computer (24) coupled to at least one acoustic sensor (22) and by a remote monitoring server (20) of an acoustic monitoring system.

Images