DEVICE FOR RECORDING A SPEECH SIGNAL OF A PERSON AND METHOD FOR RECORDING A SPEECH SIGNAL OF A PERSON WITH SUCH A DEVICE

Description

[0001] The invention relates to a detection device for a person's speech signal. The invention also relates to a motor vehicle with such a detection device and to a method for detecting a person's speech signal with such a detection device.

[0002] In the human body, the larynx, which, as part of the respiratory tract, forms the transition from the pharynx to the trachea in the front of the neck, is crucial for voice production. The laryngeal muscles vibrate the vocal cords located in the larynx to produce the voice. Various technical methods already exist for detecting and analyzing these micro-movements in the front of the neck, which occur whenever a person speaks. These methods each require a sensor positioned within direct line of sight to the person's front of the neck to capture the micro-movements that contain information about the speech signal. Therefore, an optical component is needed, preferably positioned within the person's field of vision.

[0003] US Patent 2016 / 0267911 A1 describes a vehicle speech recognition system comprising an optical sensor and a microphone. The optical sensor is located in a vehicle's rearview mirror and is designed to optically detect acoustic signals from a person inside the vehicle. The optical sensor is configured as an interferometer.

[0004] US Patent 2014 / 0119737 A1 describes a system for capturing speech-related acoustic signals using a laser-based detection device designed as a mask. This mask is worn over part of a speaker's face, covering the user's mouth. A laser microphone within the system is designed to detect vibrations of the mask and output them as speech signals.

[0005] US Patent 6,317,237 B1 describes a device that emits an infrared laser beam toward a glass pane in a building, where the beam is reflected and then captured by a receiving device. This device is intended to detect vibrations in the glass pane caused by voices in the room. In other words, it is also intended to detect and analyze speech signals within a room.

[0006] Optical diffraction gratings that are holographically fabricated and therefore referred to as holographic gratings are known from the prior art. In this regard, it is known from the scientific publication "Volume-phase holographic gratings and their potential for astronomical applications" (SC Barden, JA Arns and WS Colburn, Proceedings SPIE 3355, Optical Astronomical Instrumentation, 1998) that light striking such a holographic grating at an angle significantly outside the range of angles satisfying the Bragg condition passes through the grating undiffracted. However, if light strikes the holographic grating at an angle such that the Bragg condition is at least approximately satisfied, the light is diffracted at that angle. A similar behavior is observed with respect to the wavelength dependence of the holographic grating's influence on light.Light with a wavelength significantly outside the wavelength range defined by the Bragg condition (the so-called Bragg wavelength) passes through the holographic grating without diffracting. Only light with a wavelength that at least approximately fulfills the Bragg condition is diffracted by the holographic grating. Using complex holographic grating structures, it is therefore possible, for example, to diffract light with two different wavelength ranges at the same angle. Furthermore, a holographic grating can split light with different wavelengths into separate light paths, thus enabling the realization of a dispersive beam splitter.YEKUTIEL AVARGEL ET AL: "ROBUST SPEECH RECOGNITION USING AN AUXILIARY LASER-DOPPLER VIBROMETER SENSOR" reveals laser-based signal extraction to improve speech recognition, using a laser diode to measure laryngeal movement in order to extract improved speech features.

[0007] The object of the invention is to provide a solution by means of which a detection device for a person's speech signal can be provided that is as inconspicuous as possible.

[0008] The problem is solved by the subject matter of the independent patent claims. Advantageous embodiments of the invention are disclosed by the dependent patent claims, the following description, and the figures.

[0009] The inventive detection device for a person's speech signal comprises a light source, a sensor, an evaluation unit, and a first planar support medium. The evaluation unit, in turn, includes a conversion unit and an interpretation unit. The first light source is, for example, an emitter of a laser beam in the infrared wavelength range. The sensor is configured to detect reflected light from the light source and provide it in the form of sensor data. The sensor is, for example, a sensor configured to detect infrared light signals, i.e., an infrared sensor.

[0010] The first planar support medium is designed as an optical fiber, featuring an input area and an output area. This support medium is, for example, a transparent plastic film, and it also transmits light from the surrounding environment to the sensor. Specifically, the support medium can transmit light coupled into it to the sensor via internal reflection, preferably total internal reflection. The sensor can then detect the transmitted light once it is released from the support medium and generate sensor data from it. The light required for this process is captured or absorbed at the input and output areas of the first planar support medium. The support medium is designed as a light-conducting medium.

[0011] The light source of the detection device is designed to emit pulsed light, preferably pulsed infrared light. For this purpose, the light source can be arranged, for example, to the side of the planar support medium or next to it. The coupling area is designed as a holographic element with a first deflection structure. A description of the operation of such a holographic element, often referred to as an optical grating and which can be manufactured using holographic methods, can be found, for example, in the scientific publication cited above. The coupling area can, for example, be implemented as a diffraction grating. The first deflection structure of the coupling area is designed to couple light emitted by the light source and reflected from the person into the first planar support medium.In this example, the coupling area couples the reflected pulsed infrared light, which is reflected off the person, into the detection device itself, i.e., into the first planar carrier medium. The light source is positioned in a preferred installation location such that the pulsed light is emitted towards the front of the person's neck and reflected there. The first deflection structure is thus designed to couple the light reflected off the person into the carrier medium and deflect it sufficiently so that the coupled light fulfills a critical angle condition.

[0012] The first planar support medium is designed to transfer the coupled reflected light from the input region to the output region via internal reflection. Within the support medium, the reflected light can be guided in zigzag movements along a direction parallel to a plane on the surface of the first planar support medium. Finally, the output region, which is also designed as a holographic element, has a second deflection structure designed to couple the transferred reflected light, which falls onto the second deflection structure, out of the first planar support medium. The second deflection structure of the output region can, for example, also be implemented as a diffraction grating.

[0013] In other words, the pulsed light from the light source, reflected by the person, is deflected or diffracted at the first deflection structure of the coupling area, coupled into the carrier medium, transmitted from the carrier medium, deflected or diffracted at the second deflection structure, and coupled out of the carrier medium there. The coupling area preferably extends over the entire surface of the planar carrier medium, so that the reflected light can be captured or extracted over a relatively large area, depending on the dimensions of the planar carrier medium. Alternatively, the coupling area can extend over a partial region of the surface of the first planar carrier medium.

[0014] From the coupling area, the coupled reflected light can be transported through the first planar carrier medium to an arbitrarily chosen coupling area. To detect the light coupled out of the carrier medium, the sensor device is positioned against the coupling area. The sensor device itself is designed to detect the light coupled out in the coupling area and provide it in the form of sensor data. This sensor data describes the travel time of the pulsed light emitted by the light source and reflected by the person—that is, the light detected by the sensor device. The sensor device is therefore designed as a LIDAR device (short for light detection and ranging), since it is designed to perform a time-of-flight measurement with pulsed light. The LIDAR device is a radar-related method for optical distance and velocity measurement.In this process, the light scattered back after the emission of a light pulse is detected, and a distance to the point of scattering of the light pulses can be calculated from the travel time of the corresponding light signals.

[0015] The evaluation unit of the detection device is designed to convert the sensor data into vibration data using the conversion unit. From the time-of-flight differences of the respective pulsed light signal provided by the sensor, the distance to the person can be calculated, and thus ultimately the micro-vibrations in the front of the person's neck can be detected. This requires that the detection device is positioned in the preferred installation position so that the light from the light source is reflected in the front of the person's neck.

[0016] The evaluation unit's interpretation unit processes vibration data using a speech interpretation criterion to generate the person's speech signal. The vibration data is thus used to determine the spoken language in the form of a speech signal. This speech interpretation criterion includes, for example, vibration data correlated with the respective speech signal components that ultimately comprise the person's speech signal. The interpretation unit is designed to evaluate the vibration data, incorporating the speech interpretation criterion, in such a way that the person's speech signal can be transmitted. This process utilizes the fact that the reflected light is transmitted to the sensor device via the first carrier medium, which has a planar coupling area and an output coupling area.The first planar support medium can therefore be located, for example, in a motor vehicle, specifically, if it is transparent, in the area of ​​the vehicle's windshield. If, for example, pulsed laser light is emitted towards the driver by means of a light source positioned, say, in the area of ​​the upper dashboard, the reflected light, preferably reflected in the front neck area of ​​the person, can be captured by the planar support medium, coupled in, and transmitted to a sensor device positioned, for example, at the edge of the windshield. This allows the detection device to be integrated into the vehicle particularly discreetly, since the large surface area of ​​the windshield is used unobtrusively as a detection area for the reflected light where the transparent first planar support medium is located.

[0017] Such a recording device, which provides the person's speech signal, can be used, for example, to supplement a vehicle's speech recognition system. This speech recognition system is designed, for instance, to control a predetermined driver assistance function of the vehicle by means of recognized and interpreted acoustic commands from the driver. This offers the advantage of reducing the influence of background noise, such as engine noise or speech signals from other people in the vehicle, since ultimately the laryngeal movement, and thus the region of the speaker (i.e., the driver) relevant for generating the speech signal, is directly recorded and evaluated.

[0018] The evaluation unit, comprising the subunits of the conversion unit and the interpretation unit, includes, for example, a processor designed to provide the person's speech signal. This processor may include at least one microprocessor or at least one microcontroller. Furthermore, the processor may contain program code configured to provide the person's speech signal when executed by the processor. This program code may be stored in a data memory within the processor.

[0019] The invention also includes embodiments that offer additional advantages.

[0020] In an advantageous embodiment of the invention, it is provided that when evaluating the vibration data using the speech interpretation criterion, the provided vibration data are compared with reference vibration patterns stored in the interpretation unit, wherein the respective stored reference vibration patterns are assigned to the respective speech signals. The speech interpretation criterion thus contains respective vibration data, the so-called reference vibration patterns, which are correlated with the respective speech signal components that ultimately comprise the person's speech signal. Therefore, the interpretation unit is designed to evaluate the vibration data, with the aid of the speech interpretation criterion, in such a way that the person's speech signal can be disseminated.The speech interpretation criterion thus provides a kind of speech signal database, which allows the vibration data to be quickly and reliably assigned to corresponding speech signals.

[0021] According to the invention, the coupling area and the coupling area comprise at least one holographic surface grating or one holographic volume grating as a deflection structure. In this context, the detection device can also be referred to as HoloCam, short for holographic camera.

[0022] An optical grating, also called a diffraction grating, as well as its operating principle and manufacturing process, is, as already mentioned, generally known, as can be seen, for example, in the scientific publication cited above. In principle, an optical grating can be based on at least partially periodic structures, a so-called grating structure, embedded in a substrate. Using such a grating structure, an optical grating can achieve light redirection through the physical effect of diffraction, similar to that achieved by mirrors, lenses, or prisms. When light falls on the optical grating, that is, when light rays fall upon it, and these incident light rays, in particular, satisfy the Bragg equation, the light rays are diffracted or deflected by the optical grating.

[0023] Light guidance can thus be achieved in particular through interference phenomena of the light rays diffracted by the optical grating. The deflection structure of the input or output coupling region can accordingly also be referred to as a diffraction structure.

[0024] Preferably, an optical grating can be configured to be directionally or angle-selective with respect to the incident light. Thus, only light, and in particular a portion of the light, that falls onto the optical grating from a predetermined direction of incidence, for example at a predetermined angle, can be deflected. Light, and in particular a portion of the light, that falls onto the optical grating from a different direction is preferably not deflected, or the deflection decreases the greater the difference from the predetermined direction of incidence. The portion of light that deviates from the predetermined or optimal direction of incidence can therefore preferably propagate unhindered through the substrate containing the optical grating.

[0025] Additionally or alternatively, an optical grating can be wavelength-selective or frequency-selective. Thus, only light, in particular a first component of the light with a predetermined wavelength, can be deflected or diffracted by the optical grating at a specific diffraction angle. Light, in particular a second component of the light with a wavelength other than the predetermined one, is preferably not deflected, or the deflection decreases the greater the difference from the predetermined wavelength. The second component of light, which deviates from the predetermined wavelength or optimal wavelength, can therefore preferably propagate unhindered through the substrate containing the optical grating. This allows, for example, at least a monochromatic component of light to be separated from polychromatic light striking the optical grating. In Advantageously, the deflection effect is maximal for the optimal wavelength and decreases or weakens towards longer and shorter wavelengths, for example according to a Gaussian curve. In particular, the deflection effect only affects a fraction of the visible light spectrum and / or within an angular range of less than 90 degrees.

[0026] Optical gratings can be fabricated, in particular, by exposing a substrate, for example, photolithography or holography. In this context, the optical grating can also be referred to as a holographic or holographic-optical grating. Two types of holographic-optical gratings are known: surface holographic gratings (SHGs) and volume holographic gratings (VHGs). In a surface holographic grating, the grating structure can be created by optically deforming the surface structure of the substrate. The altered surface structure allows incident light to be deflected, for example, reflected. Examples of holographic surface gratings are sawtooth and blaze gratings.In contrast, the grating structure of holographic volume gratings can be incorporated into the entire volume or a portion of the substrate's volume. Holographic surface gratings and holographic volume gratings are generally frequency-selective. However, optical gratings capable of diffracting polychromatic light are also known. These are called multiple-volume holographic gratings (MVHGs) and can be produced, for example, by modifying the periodicity of an optical grating's structure or by arranging several holographic volume gratings in series.

[0027] Suitable materials for the substrate used to incorporate an optical grating include a polymer, especially a photopolymer, or a film, especially a photosensitive film, for example made of plastic or organic materials. Substrates that have a deflection structure for diffracting light, for example in the form of an optical grating, can also be called holographic optical elements (HOEs).

[0028] The described design of the coupling area and the coupling area therefore makes it possible to diffract the light falling on the coupling area to the sensor device arranged, for example, laterally on the cover plate, which ultimately allows the speech signal to be provided after the evaluation steps described above.

[0029] Another particularly advantageous embodiment of the invention provides that a light coupling area and a light coupling area are provided on a second planar support medium. The detection device can therefore comprise a second support medium in addition to the first. The light coupling area is designed as a holographic element with a third deflection structure. The third deflection structure is designed to couple light falling from the light source of the detection device onto the third deflection structure into the second planar support medium. The third deflection structure of the light coupling area can, for example, be implemented as a diffraction grating. The light coupling area is positioned such that the pulsed light emitted by the light source can be introduced into the second planar support medium via the light coupling area.The second planar support medium is designed to transfer the coupled light from the light coupling area to the light coupling area via internal reflection. The light coupling area is itself designed as a holographic element with a fourth deflection structure, which is, for example, also implemented as a diffraction grating. This fourth deflection structure is designed to couple the transmitted light, which falls upon it, out of the second support medium and emit it towards the person.It is therefore possible that, via a planar carrier medium, described here as the second planar carrier medium, the light from the light source is transported from the light source to the light output coupling area, from which the pulsed light is emitted towards the person, whereupon it is reflected at the person and transmitted via the coupling area into the first planar carrier medium and ultimately to the sensor device.

[0030] The second planar support medium can be separate from the first planar support medium and, for example, can also be designed as a transparent film. It is possible, for instance, for the light source to be located in the area of ​​the first A-pillar of a motor vehicle. The light from the light source is then transmitted via the second planar support medium, which is located on the windshield of the motor vehicle, to a central point in front of the driver's seat and emitted towards the front of the driver's neck by means of a light-emitting area positioned there. In this example, the first support medium, which is also a transparent film on the windshield, can be positioned directly next to the second planar support medium, so that the input area is located close to the light-emitting area or even at least partially overlaps with it.At the coupling point, the light reflected from the front of the person's neck is coupled into the first carrier medium, which then transmits the light, for example, to a second A-pillar of the vehicle, where the sensor device and, for example, the evaluation unit are located. This makes it possible to implement a completely inconspicuous detection device that is invisible to the driver of the vehicle in the described installation position, especially if the light source emits light in a non-visible wavelength range, such as infrared light.By exploiting the properties of the HOE, it is therefore possible to ensure that neither the light source nor the sensor device needs to be located within the field of vision of the person whose speech signal is to be provided, but rather that a signal transmission of the light emitted by the light source and the light reflected by the person is possible via the HOE to the respective predetermined target position, i.e. to the front neck region of the person or to the sensor device.

[0031] According to a particularly advantageous embodiment of the invention, the first planar support medium and the second planar support medium are designed as a common main support medium, on which the coupling area, the coupling area, the light coupling area, and the light coupling area are provided. Thus, the detection device comprises only one support medium, the so-called main support medium; that is, the first planar support medium and the second planar support medium are spatially connected. The planar design of the common main support medium therefore provides a large light emission area for the light from the light source and a large light entry area for the reflected light, thereby compensating for measurement inaccuracies to a certain extent.For example, it becomes possible for a wavefront to be emitted towards the person via the HOE, and for the reflected wavefront to be re-coupled via the predetermined light entry area of ​​the carrier medium's coupling region and transmitted to the sensor device. This allows for various configurations of the emitted pulsed light and, consequently, different designs of the detection device. Ultimately, this ensures a reliable transmission path for the person's speech signal, as it provides a reliable transmission path for the light from the light source to the front of the person's neck and for the reflected light from the front of the person's neck to the sensor device.

[0032] In a further embodiment of the invention, the detection device includes a microphone. The detection device can thus include a transducer that converts airborne sound as alternating pressure vibrations into corresponding electrical voltage changes, i.e., corresponding microphone signals. Such a microphone may, for example, already be provided in a motor vehicle, for instance, for a speech recognition system for voice control of vehicle functions. The microphone is designed to capture the person's speech data and provide it to the evaluation unit. The evaluation unit, in turn, is designed to provide output data for the interpretation unit of the evaluation unit by applying a weighting criterion to the provided vibration and speech data.A comparison is therefore made between the sensor data acquired by the microphone and the sensor data provided to the detection device by the sensor unit, as described above, from which the conversion unit has determined the vibration data. The weighting criterion can, for example, determine, depending on the current driving condition and / or a frequency range of the speech data, whether the speech data correlates with the driver speaking. If necessary, the weighting criterion can also prevent background noise from having a strong influence on the output data and thus ultimately on the quality of the provided speech signal. The output data provided to the interpretation unit is evaluated in such a way that, applying the speech interpretation criterion, the speech signal of the person is provided.The vibration data provided by the conversion unit is ultimately used, in conjunction with the microphone's speech data, to analyze the noise in the vehicle and to extract and process the speech contained therein.

[0033] For example, the microphone's speech data can be optimized using vibration data and a weighting criterion to provide the purest possible speech signal with minimal background noise. This ultimately optimizes the person's speech signal, ensuring it accurately reflects their actual spoken language.

[0034] The invention further provides that the detection device comprises an image acquisition unit. The image acquisition unit is, for example, configured as an image sensor or camera sensor, such as a CCD sensor (charged-coupled device) or a CMOS sensor (complementary metal oxide semiconductor). The image acquisition unit is configured to detect ambient light incident on the coupling area from the environment, transmit it from the first planar support medium to the output area, and extract it at the output area. This data is then provided to the evaluation unit in the form of image data that correlates with the detected ambient light.In such a configuration of the image acquisition device as an image sensor, the first planar support medium, on which the coupling and output areas are arranged, can additionally assume the function of a lens, i.e., an imaging optic. Alternatively, the image acquisition device can be designed as a camera or photographic apparatus, in particular as a microcamera, and consequently be implemented with its own imaging optic. The image data, which correlates with the captured environmental image, contains, for example, image information about the driver, who, in the preferred installation position described above, is sitting in front of the acquisition device in the vehicle.

[0035] The evaluation unit is designed to provide output data for the interpretation unit by applying a comparison criterion to the supplied vibration and image data. The image data contains, for example, information about the current orientation of the driver's neck region or head relative to a predefined normal orientation, in which the driver is looking straight ahead through the windshield in the longitudinal direction of the vehicle. If the driver moves their head forward or backward relative to the windshield, or tilts it relative to the described normal orientation, this can be detected using the image data.Any resulting effect on the vibration data can thus be corrected using the adjustment criterion so that the output data reliably describes the actual vibration of the person's front neck region and is not influenced, for example, by the driver's slow movements relative to these vibrations. The interpretation unit is then designed to provide the person's speech signal by evaluating the output data and applying the speech interpretation criterion. The driver's movement, independent of any movement in the front neck region while speaking, can therefore be taken into account, ensuring that the person's speech signal is always reliably provided without the quality of the speech signal being affected by relative movements of the driver's head and neck area relative to the detection device.

[0036] According to the invention, the planar support medium between the coupling area and the coupling area is designed as a transparent plate, film, or lacquer. The planar support medium can therefore be of small thickness, such that its width and length are large compared to this small thickness, which is perpendicular to the planar surface of the support medium. The planar support medium can, for example, be between half a millimeter and five millimeters thick. If the planar support medium is designed as a transparent film, it is also flexible, meaning it can be deformed non-destructively. Non-destructive deformation is defined as bending the film around a bending radius of less than two centimeters.If the flat carrier medium is designed as a transparent coating, it can have a thickness in the micrometer range and consequently less than one millimeter. This allows the flat carrier medium, with its coupling and release areas, to be discreetly positioned in numerous locations, for example, inside a vehicle or on a wall, without obscuring any objects behind it. This makes placement on a vehicle window, for instance, conceivable. As a result, the detection device can be integrated into any environment.

[0037] In a further embodiment of the invention, the coupling area and the decoupling area are formed integrally with the carrier medium, or the carrier medium is formed as a separate element from the coupling area and the decoupling area. In the first case, the coupling area and the decoupling area can, for example, be incorporated directly onto a surface of the carrier medium. That is, the deflection structure can, for example, be etched or laser-etched into the surface of the carrier medium. Thus, the carrier medium itself can be designed as a high-efficiency element (HOE). In the second case, the coupling area, the decoupling area, and the carrier medium can be formed separately. In this case, the coupling area and the decoupling area can, for example, form at least one first element, and the carrier medium can form a second element that abuts the first element.Thus, the coupling area and the output area can be formed in at least one HOE (holographic optical element). For example, the coupling area and the output area can be formed in different sections of a holographic film or plate. To attach the film or plate to the substrate, the film can be bonded to the substrate. Alternatively, the holographic film can also be designed as an adhesive film and adhere directly to the surface of the substrate by molecular forces, i.e., without adhesive. The detection device with the substrate and the coupling and output areas arranged thereon can therefore be manufactured in various ways and, in particular, cost-effectively.

[0038] The motor vehicle according to the invention comprises a detection device as described above. The first planar carrier medium, with its coupling area and coupling area, is arranged in one of the following positions in the motor vehicle: in a screen of a display device in the vehicle interior, in a rearview mirror, in a center console, in a dashboard, in a combination instrument, in a windshield, in a side window, and / or between two A-pillars of the motor vehicle. The first planar carrier medium can be arranged on a surface of the aforementioned component of the motor vehicle, particularly if the first planar carrier medium is designed as a transparent carrier medium, for example, as a film, cover plate, or paint.This allows the detection device to be used for measuring in at least two directions: inwards, towards the interior of the vehicle, for example towards the driver, and outwards towards the vehicle's surroundings. This means, for example, that the speech signal of a person standing outside the vehicle, whose vibrations in the front of their neck are detected by the device and made available to them as a speech signal, can be output, for example, via an output device inside the vehicle. This allows the driver to hear what the person outside the vehicle is saying at any given moment without having to open a window.For this purpose, the first planar support medium can, for example, be located on the outer surface of the windshield or at least one of the side windows of the vehicle. Ultimately, the only relevant factor is that the first planar support medium is positioned inside or on the outer wall of the vehicle in such a way that the light from the light source can be emitted in such a way that it strikes the person, particularly the front neck region, and that the reflected light also strikes the coupling area of ​​the first planar support medium, so that the detection device is ultimately arranged in the preferred installation position described above. Ultimately, the detection device can be used in various positions within a vehicle and for the various purposes mentioned.

[0039] The invention also includes embodiments of the motor vehicle according to the invention that have features already described in connection with the further developments of the detection device according to the invention. For this reason, the corresponding further developments of the motor vehicle according to the invention are not described again here.

[0040] The motor vehicle according to the invention is preferably designed as a motor vehicle, in particular as a passenger car or truck, or as a passenger bus or motorcycle.

[0041] In an advantageous embodiment of the motor vehicle according to the invention, the vehicle comprises at least one detection device, in particular a camera device, an acceleration detection device, and / or a gyroscope. A gyroscope is also sometimes referred to as a gyroscope. The at least one detection device is configured to provide the data it acquires to the evaluation unit of the detection device. The evaluation unit of the detection device is then designed to process the data from the vehicle's detection device to generate the person's voice signal using the interpretation unit. The detection device can also acquire further parameters, such as those describing the current movement of the vehicle.

[0042] For example, during a vehicle journey, the driver may, due to inertial forces, perform a certain relative movement to the vehicle's frame of reference. This movement can ultimately influence the sensor data and thus the vibration data on which the speech signal is based. Therefore, the detection system uses additional data regarding the vehicle's movement and / or the movement of the occupants, such as that obtained from the camera system. This data is then used to reduce potential sources of error when generating the speech signals. The advantage of this is that a reliable speech signal from the person can be evaluated.Information about the movement of the motor vehicle itself, and in particular about inertial forces within the motor vehicle, can be recorded using the acceleration detection device and / or the yaw rate detection device, whereby the yaw rate detection device is used to record the rotational movement of the motor vehicle, for example, around its own axis.

[0043] The method according to the invention is designed to capture a speech signal from a person using a detection device, wherein the detection device comprises a light source, a sensor device, an evaluation unit comprising a conversion unit and an interpretation unit, and a first planar support medium. The first planar support medium is designed as an optical fiber, on which an input area, configured as a holographic element with a first deflection structure, and an output area, configured as a holographic element with a second deflection structure, are provided. The first planar support medium with the input area and the output area is configured as a transparent plate, film, or lacquer, and the input area and the output area each have at least one holographic volume lattice or a holographic surface lattice as their respective deflection structures.

[0044] The method according to the invention comprises the following steps: Emitting pulsed light by means of the light source in a direction towards the front of the person's neck; coupling in light emitted by the light source and reflected at the front of the person's neck by means of the first deflection structure; transferring the coupled reflected light by means of internal reflection from the coupling area to the coupling area by means of the first planar carrier medium; coupling out the transferred reflected light, which falls onto the second deflection structure, from the first planar carrier medium by means of the second deflection structure; detecting the light coupled out in the coupling area and providing sensor data, which describe a transit time of the pulsed light emitted by the light source and reflected at the front of the person's neck, by means of the sensor device;Converting the sensor data into vibration data describing a relative movement of the person's anterior neck region using the conversion unit of the evaluation device, and providing the person's speech signal using the interpretation unit of the evaluation device by evaluating the vibration data by applying a speech interpretation criterion to the vibration data.

[0045] The detection device comprises an image capture unit that captures ambient light incident on the input area from the surroundings. This light is transmitted from the first planar carrier medium to the output area, where it is extracted and provided to the evaluation unit in the form of image data that correlates with the captured ambient light. Applying a matching criterion to the provided vibration and image data, the evaluation unit provides output data for the interpretation unit. By evaluating this output data using the speech interpretation criterion, the person's speech signal is derived.

[0046] The preferred embodiments and their advantages presented in connection with the detection device according to the invention apply accordingly, insofar as applicable, to the method according to the invention. For this reason, the corresponding further developments of the method according to the invention are not described again here.

[0047] The invention also includes a control device for the detection device, wherein the control device is designed to carry out the method according to the invention, which is performed by means of the detection device. This control device comprises a processor unit configured to perform an embodiment of the method according to the invention. For this purpose, the processor unit can comprise at least one microprocessor and / or at least one microcontroller. Furthermore, the processor unit can comprise program code configured to perform the embodiment of the method according to the invention when executed by the processor unit. The program code can be stored in a data memory of the processor unit.

[0048] The invention also includes combinations of the features of the described embodiments.

[0049] The following are exemplary embodiments of the invention described. This is illustrated by: Fig. 1 a schematic representation of a motor vehicle with a detection device; Fig. 2 a schematic representation of a detection device with a first planar carrier medium; and Fig. 3 a schematic representation of a detection device with a common main carrier medium.

[0050] The exemplary embodiments described below are preferred embodiments of the invention. In these exemplary embodiments, the described components each represent individual features of the invention, which can be considered independently of one another and each further develops the invention independently. Therefore, the disclosure is intended to include combinations of features of the embodiments other than those shown. Furthermore, the described embodiments can also be supplemented by further features of the invention already described.

[0051] In In the figures, the same reference symbols denote functionally equivalent elements.

[0052] In Fig. 1 A detection device 10 for a speech signal from a person 31 is sketched. This detection device 10 is arranged in a motor vehicle 32, specifically in a vehicle interior 34 of the motor vehicle 32.

[0053] The detection device 10 comprises a light source 30 designed as an infrared light source. The light source 30 is designed to emit pulsed light 100. The detection device 10 also comprises a sensor device 11 designed as a LiDAR sensor. The detection device 10 further comprises an evaluation device 13. The detection device 10 also comprises a first planar support medium 12 designed as a light guide. In addition, the detection device 10 comprises a second planar support medium 12', which is positioned laterally to the first planar support medium 12 in an upper area of ​​a windshield 36 of the motor vehicle 32. The two planar support media 12, 12' are thus arranged between two A-pillars 37 of the motor vehicle 32.

[0054] Both the first planar support medium 12 and the second planar support medium 12' can each be designed as a transparent plate, film, or lacquer. In this example, they are each designed as a transparent film, each arranged on an inner surface of the windshield 36. Fig. 1 It becomes clear that pulsed light 100 emitted by the light source 30 is coupled into the second planar carrier medium 12' in a light coupling area 17 and emitted in a light coupling area 19 of the second planar carrier medium 12' in such a way that it strikes a front neck region 31' of the person 31. There, this pulsed light 100 is reflected, so that the reflected light 100' is emitted back towards the windshield 36. There, in a coupling area 16, the reflected light 100' is coupled into the first planar carrier medium 12 and transferred to an output coupling area 18, to which the sensor device 11 is located. The coupling area 16 and the coupling area 18 are therefore provided on the first planar support medium 12, whereas the light coupling area 17 and the light coupling area 19 are provided on the second planar support medium 12'.

[0055] The detection device 10 also comprises an evaluation unit 13, a microphone unit 40, an image capture unit 42, and a detection unit 44, wherein, in this example, the detection unit 44 is an acceleration detection unit that detects the acceleration of the motor vehicle 32. Alternatively, the detection unit 44 could be a camera unit and / or a gyroscope. The detection unit 44 is thus designed to detect the movement of the motor vehicle 32 and / or the person 31 in the motor vehicle 32 and to provide it in the form of corresponding detection data.

[0056] In Fig. 2 The diagram details how the speech signal of person 31 can be provided using the detection device 10. It clearly shows that the coupling area 16 is designed as a holographic element 14 with a first deflection structure 20. The first deflection structure 20 is designed to couple the light 100', which is emitted by the light source 30 and reflected from person 31, specifically from the anterior neck region 31' of person 31, into the first planar carrier medium 12. The first planar carrier medium 12 is then designed to transfer the coupled reflected light 100' from the coupling area 16 to the output coupling area 18 by means of internal reflection. The coupling area 16 is also designed as a holographic element 14, but with a second deflection structure 22.The second deflection structure 22 is designed to couple the transmitted reflected light 100', which falls onto the second deflection structure 22, out of the first planar support medium 12. The coupling area 16 and the coupling area 18 each have, as deflection structures 20 and 22, at least one optical grating, in particular a holographic volume grating or a holographic surface grating.

[0057] The sensor device 11 is configured to detect the light 100' extracted in the extraction area 18 and to provide it in the form of sensor data. The sensor data describes the travel time of the pulsed light 100' emitted by a light source 30 and reflected by the person 31, which was detected by the sensor device 11. The sensor device 11 is therefore a type of LiDAR device that can detect and provide the travel time of the light 100'. The evaluation device 13 comprises a conversion unit 13a and an interpretation unit 13b. The evaluation device 13 is configured to convert the sensor data into vibration data using the conversion unit 13a and to provide the speech signal of the person 31 by evaluating the vibration data using a speech interpretation criterion using the interpretation unit 13b.When evaluating the vibration data using the speech interpretation criterion, the provided vibration data are compared with reference vibration patterns stored in the interpretation unit 13b, whereby the respective stored reference vibration patterns are assigned to the respective speech signals.

[0058] To capture the speech signal of person 31, the following steps are taken: In step S1, pulsed light 100 is emitted by the light source 30. In step S2, the light 100' emitted by the light source 30 and reflected by person 31 is coupled into the first planar carrier medium 12 by means of the first deflection structure 20. Subsequently, in step S3, the coupled reflected light 100' is transferred by internal reflection from the input area 16 to the output area 18 by means of the first planar carrier medium 12. Subsequently, in step S4, the transferred reflected light 100', which falls on the second deflection structure 22, is coupled out of the first planar carrier medium 12 by means of the second deflection structure 22.Subsequently, the light 100' coupled out in the coupling area 18 is detected by the sensor device 11 in step S5, whereupon, in step S6, sensor data describing the transit time of the pulsed light 100' emitted by the light source 30 and reflected by the person 31 are provided by the sensor device 11. In a further step S7, the sensor data are converted into vibration data by means of the conversion unit 13a of the evaluation unit 13. In In step S8, the speech signal of person 31 is finally provided by the interpretation unit 13b by evaluating the vibration data using the speech interpretation criterion. It is taken into account that the light source 30 emits the light 100 in such a way that it is directed towards the anterior neck region 31' of person 31 and reflected there. The reflected light 100' thus contains information about vibrations in a laryngeal region of person 31, as would be expected in a speaking person 31, and can therefore be measured and evaluated.

[0059] In Fig. 3 Instead of the first planar support medium 12, a common main support medium 12" is sketched, which encompasses both the first planar support medium 12 and the second planar support medium 12'. The coupling area 16 and the coupling area 18, as well as the light coupling area 17 and the light coupling area 19, are therefore provided on the common main support medium 12'. Fig. 1 The first planar support medium 12 and the second planar support medium 12' were shown separately, so that the coupling area 16 and the coupling area 18 are provided on the first planar support medium 12 and the light coupling area 17 and the light coupling area 19 are provided on the second planar support medium 12'.

[0060] The light coupling area 17 is designed as a holographic element 14 with a third deflection structure 24, as shown in Fig. 3 The third deflection structure 24 is designed to couple light 100, which falls onto the third deflection structure 24 from the light source 30 of the detection device 10, into the main carrier medium 12". If, alternatively, a separate second planar carrier medium 12' is encompassed by the detection device 10 instead of the main carrier medium 12", the light 100 from the light source 30 is coupled into the second planar carrier medium 12' by the third deflection structure 24. The main carrier medium 12" is then designed to transfer the coupled light 100 from the light coupling area 17 to the light coupling area 19 by means of internal reflection. The light coupling area 19 is also designed as a holographic optical element 14 with a fourth deflection structure 26.The fourth deflection structure 26 is designed to couple the transmitted light 100 that falls upon the fourth deflection structure 26 from the main carrier medium 12" (or from the second carrier medium 12') and emit it in the direction of person 31, in particular the anterior neck region 31' of person 31. The path of the reflected light 100' then proceeds analogously as in . Fig. 2 described.

[0061] The microphone device 40 of the detection device 10, which is in Fig. 1 The device outlined is designed to capture speech data from person 31 and provide it to the evaluation unit 13. The evaluation unit 13 is then designed to apply a weighting criterion to the provided vibration and speech data, thereby providing output data for the interpretation unit 13b and, by evaluating this data or applying the speech interpretation criterion, distributing the speech signal of person 31.

[0062] The image acquisition device 42 of the acquisition device 10, which is also sketched in Fig., is designed to capture ambient light that strikes the coupling area 16 from the environment, transmit it through the first planar carrier medium 12 to the coupling area 18, and couple it out at the coupling area 18. This light is then provided to the evaluation unit 13 in the form of image data that correlates with the captured ambient light. The evaluation unit 13 is designed to provide output data for the interpretation unit 13b by applying a matching criterion to the provided vibration data and image data. By evaluating this output data using the speech interpretation criterion, the speech signal of person 31 is then provided.

[0063] The also in Fig. 1 The sketched detection device 44, that is, in this example the acceleration detection device, is designed to provide the data it acquires to the evaluation device 13. The evaluation device 13 is designed to evaluate the acquisition data to generate the speech signal of person 31 using the interpretation unit 13b. For example, the acceleration of the vehicle 32 can be taken into account, as this causes person 31 to perform a predetermined relative movement to the vehicle 32 in the front neck region 31', which may influence the sensor data provided by the sensor device 11. By considering this acceleration data of the vehicle 32, possible errors in the determined vibration data can be corrected, ensuring that the quality of person 31's speech signal remains optimal even while the vehicle 32 is in motion.

[0064] The light source 30, the evaluation unit 13, the sensor unit 11 and the image acquisition unit 42 are in the in Fig. 1 The sketched motor vehicle 32 is positioned in one of the two A-pillars 37 of the motor vehicle 32.

[0065] As an alternative to the described position of the first planar carrier medium 12 and the second planar carrier medium 12' in the windshield 36 of the motor vehicle 32, the first planar carrier medium 12 and / or the second planar carrier medium 12' and / or the common main carrier medium 12" can be positioned in a screen of a display device in the vehicle interior 34 of the motor vehicle 32, in a rearview mirror, in a center console, in a dashboard, in a combination instrument and / or in a side window.

[0066] Overall, the examples demonstrate how the invention can be used to provide laser interferometry, for example, for in-vehicle telephone calls. This provides a contactless throat microphone for a motor vehicle 32. Ultimately, this enables an optical rangefinder to be integrated into a window pane of the motor vehicle 32 or another position within the vehicle 32 using the first planar support medium 12 with the coupling area 16 and the coupling area 18, i.e., a holographic optical element (HOE). The described exemplary integration into a window pane of the motor vehicle 32, specifically the windshield 36, reduces the risk of the detection area being obstructed, for example, by the driver's hand.The distance to the object being measured, the anterior neck region 31' of person 31, can also be reduced, for example, by positioning at least the first planar carrier medium 12 in a side window of the motor vehicle 32. Furthermore, the detection device 10 can be used in two directions: inwards towards the vehicle interior 34 and outwards towards the surroundings of the motor vehicle 32. This allows, for example, the voice signal from person 31 standing outside the motor vehicle 32 to be transmitted into the vehicle interior 34 without having to open the corresponding window.

[0067] The technical implementation is achieved by means of an emitter, that is, the light source 30, which emits pulsed light 100, for example, a laser beam as a measuring beam. The emitted light 100 is coupled into an optical fiber made of a transparent substrate material, the first planar carrier medium 12. The first planar carrier medium 12 can be positioned, for example, on a window pane of the vehicle 32, but also, for example, on the screen of a mobile device or on another device. The emitted light 100 is thus coupled in via a holographic function, that is, via the second planar carrier medium 12' or the common main carrier medium 12", and guided to a respective destination in order to be coupled out there via the light output coupling area 19 in the direction of the person 31.The coupling of the light 100 into and out of the respective carrier medium 12, 12' can take place on the respective carrier medium or at its edge. However, it can also be coupled out in the middle of the person 31's field of vision, in their direction, where the installation of the light source 30 and the sensor device 11 would not be possible due to its physical presence, as it cannot, for example, be positioned centrally on the windshield 36, since the windshield must be transparent so that the person 31 can see through it. The detection area, i.e., the coupling area 16, is preferably designed as a flat surface, which ultimately increases the sensor area of ​​the sensor device 11. This allows inaccuracies in the measurement to be compensated for to a certain extent.The light 100' reflected from the anterior neck region 31' of person 31 can be received via the same main carrier medium 12" or via the first planar carrier medium 12 through the coupling area 16 and transmitted to the sensor device 11. The sensor device 11 can thus be arranged, like the light source 30, offset from the actual point where the light 100 is emitted towards person 31 and the reflected light 100' is coupled into the HOE.

[0068] Using the sensor data acquired by sensor device 11, the distance to person 31 can now be calculated and thus the vibration of the anterior neck region 31' can be determined. The vibration patterns obtained in this way are then used, for example, to analyze the sounds acquired by microphone device 40, to extract and process the speech contained therein. Thus, based on the vibration data, the speech signal of person 31 is processed and made available by the interpretation unit 13b by evaluating the vibration data and applying a predefined speech interpretation criterion.

[0069] Since the motor vehicle 32 is a moving system and inertial effects can cause the driver, i.e., person 31, to behave differently than a rigid body of the motor vehicle 32 around them, additional information about the movement of the motor vehicle 32, i.e., the movement of person 31 within the motor vehicle 32, can be used. This information is obtained by means of the detection device 44, that is, by means of the camera device of the acceleration detection device and / or the yaw rate detection device. Based on this data from the detection device 44, the determination of the speech signal of person 31 can be improved, and sources of error for faulty speech signals from person 31 can also be reduced.

Claims

1. A detection device (10) for a voice signal of a person (31), wherein the detection device (10) includes a light source (30), a sensor device (11), an evaluation device (13), which comprises a conversion unit (13a) and an interpretation unit (13b), as well as a first planar carrier medium (12), wherein the first planar carrier medium (12) is formed as a light guide, at which a coupling-in area (16) and a coupling-out area (18) are provided, wherein the first planar carrier medium (12) with the coupling-in area (16) and the coupling-out area (18) is formed as a transparent plate, foil or varnish, wherein - the light source (30) is formed to emit pulsed light (100); - the coupling-in area (16) is formed as a holographic element (14) with a first deflection structure (20), which is configured to couple light (100'), which is emitted from the light source (30) and reflected on the person (31), into the first planar carrier medium (12); - the first planar carrier medium (12) is formed to transfer the coupled-in reflected light (100') from the coupling-in area (16) to the coupling-out area (18) by means of internal reflection; - the coupling-out area (18) is formed as a holographic element (14) with a second deflection structure (22), which is configured to couple the transferred reflected light (100'), which is incident on the second deflection structure (22), out of the first planar carrier medium (12), wherein the coupling-in area (16) and the coupling-out area (18) comprise at least one holographic volume grating or holographic surface grating as the respective deflection structure (20, 22); - the sensor device (11) is formed to detect the light (100') coupled out in the coupling-out area (18) and to provide it in the form of sensor data, which describes a travel time duration of the pulsed light (100') emitted from the light source (30) and reflected on the person (31); and - the evaluation device (13) is formed to convert the sensor data into oscillation data by means of the conversion unit (13a) and to provide the voice signal of the person (31) by means of the interpretation unit (13b) by evaluating the oscillation data by applying a voice interpretation criterion, wherein the voice interpretation criterion contains respective oscillation data, which is correlated with respective voice signal components, from which the voice signal of the person is composed, such that the interpretation unit (13b) is configured to evaluate the oscillation data by consulting the voice interpretation criterion such that the voice signal of the person (31) can be provided, wherein the detection device (10) includes an image detection device (42), which is formed to detect ambient light, which impinges from the environment onto the coupling-in area (16), is transferred to the coupling-out area (18) by the first planar carrier medium (12) and is coupled out at the coupling-out area (18), and to provide it in the form of image data, which correlates with the detected ambient light, to the evaluation device (13), wherein the evaluation device (13) is configured to provide output data for the interpretation unit (13b) by applying a comparison criterion to the provided oscillation data and image data and to provide, by evaluation thereof by applying the voice interpretation criterion, the voice signal of the person (31).

2. The detection device (10) according to the preceding claim, wherein in evaluating the oscillation data by applying the voice interpretation criterion, the provided oscillation data is compared to reference oscillation patterns recorded in the interpretation unit (13b), wherein the respective recorded reference oscillation patterns are associated with respective voice signals.

3. The detection device (10) according to any one of the preceding claims, wherein a light coupling-in area (17) and a light coupling-out area (19) are provided at a second planar carrier medium (12'), wherein the light coupling-in area (17) is formed as a holographic element (14) with a third deflection structure (24), which is configured to couple light (100), which is incident from the light source (30) of the detection device (10) to the third deflection structure (24), into the second planar carrier medium (12'); the second planar carrier medium (12') is formed to transfer the coupled-in light (100) from the light coupling-in area (17) to the light coupling-out area (19) by means of internal reflection; and the light coupling-out area (19) is formed as a holographic element (14) with a fourth deflection structure (26), which is configured to couple the transferred light (100), which is incident on the fourth deflection structure (26), out of the second carrier medium (12') and to emit it towards the person (31).

4. The detection device (10) according to the preceding claim, wherein the first planar carrier medium (12) and the second planar carrier medium (12') are formed as a common main carrier medium (12"), at which the coupling-in area (16), the coupling-out area (18), the light coupling-in area (17) and the light coupling-out area (19) are provided.

5. The detection device (10) according to any one of the preceding claims, wherein the detection device (10) includes a microphone device (40), which is configured to detect voice data of the person (31) and to provide it to the evaluation device (13), wherein the evaluation device (13) is configured to provide output data for the interpretation unit (13b) by applying a weighting criterion to the provided oscillation data and voice data and to provide, by evaluation thereof by applying the voice interpretation criterion, the voice signal of the person (31).

6. A motor vehicle (32) with a detection device (10) according to any one of the preceding claims, wherein the first planar carrier medium (12) with the coupling-in area (16) and the coupling-out area (18) is arranged in one of the following positions: - in a screen of a display device in the vehicle interior (34); - in a rearview mirror; - in a center console; - in a dashboard; - in an instrument cluster; - in a windshield (36); - in a side window; - between two A-pillars (37) of the motor vehicle (32).

7. The motor vehicle (32) according to the preceding claim, wherein the motor vehicle (32) includes at least one detection device (44), in particular a camera device, an acceleration detection device and / or a rotation rate detection device, wherein the at least one detection device (44) is formed to detect additional data relating to a movement of the motor vehicle (32) and / or a movement of occupants of the motor vehicle (32) and to provide this detection data detected by it for the evaluation device (13), and the evaluation device (13) is configured to evaluate the detection data for generating the voice signal of the person (31) by means of the interpretation unit (13b).

8. A method for detecting a voice signal of a person (31) with a detection device (10), which includes a light source (30), a sensor device (11), an evaluation device (13), which comprises a conversion unit (13a) and an interpretation unit (13b), as well as a first planar carrier medium (12), wherein the first planar carrier medium (12) is formed as a light guide, at which a coupling-in area (16), which is formed as a holographic element (14) with a first deflection structure (20), and a coupling-out area (18), which is formed as a holographic element (14) with a second deflection structure (22), are provided, wherein the first planar carrier medium (12) with the coupling-in area (16) and the coupling-out area (18) is formed as a transparent plate, foil or varnish, and the coupling-in area (16) and the coupling-out area (18) comprise at least one holographic volume grating or holographic surface grating as the respective deflection structure (20, 22), comprising the steps: - emitting pulsed light (100) by means of the light source (30) towards a front neck region (31') of the person (31) (S1); - coupling-in light (100'), which is emitted from the light source (30) and reflected on the front neck region (31') of the person (31), by means of the first deflection structure (20) (S2); - transferring the coupled-in reflected light (100') by means of internal reflection from the coupling-in area (16) to the coupling-out area (18) by means of the first planar carrier medium (12) (S3); - coupling the transferred reflected light (100'), which is incident on the second deflection structure (22), out of the first planar carrier medium (12) by means of the second deflection structure (22) (S4); - detecting the light (100') coupled out in the coupling-out area (18) (S5) and providing sensor data, which describes a travel time duration of the pulsed light (100') emitted from the light source (30) and reflected on the front neck region (31') of the person (31), by means of the sensor device (11) (S6); - converting the sensor data into oscillation data, which describes a relative movement of the front neck region (31') of the person (31), by means of the conversion unit (13a) of the evaluation device (13) (S7) and providing the voice signal of the person (31) by means of the interpretation unit (13b) of the evaluation device (13) by evaluating the oscillation data by applying a voice interpretation criterion to the oscillation data (S8), wherein the voice interpretation criterion contains respective oscillation data, which is correlated with respective voice signal components, from which the voice signal of the person is composed, such that the interpretation unit (13b) is configured to evaluate the oscillation data by consulting the voice interpretation criterion such that the voice signal of the person (31) can be provided, wherein the detection device (10) includes an image detection device (42), which detects ambient light, which impinges from the environment onto the coupling-in area (16), is transferred to the coupling-out area (18) by the first planar carrier medium (12) and is coupled out at the coupling-out area (18), and provides it in the form of image data, which correlates with the detected ambient light, to the evaluation device (13), wherein the evaluation device (13) provides output data for the interpretation unit (13b) by applying a comparison criterion to the provided oscillation data and image data and provides, by evaluation thereof by applying the voice interpretation criterion, the voice signal of the person (31).