WO2023198393A1 - Procédé pour commander un véhicule - Google Patents

Procédé pour commander un véhicule Download PDF

Info

Publication number
WO2023198393A1
WO2023198393A1 PCT/EP2023/056645 EP2023056645W WO2023198393A1 WO 2023198393 A1 WO2023198393 A1 WO 2023198393A1 EP 2023056645 W EP2023056645 W EP 2023056645W WO 2023198393 A1 WO2023198393 A1 WO 2023198393A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
acoustic
image stabilization
acceleration sensor
natural frequency
Prior art date
Application number
PCT/EP2023/056645
Other languages
German (de)
English (en)
Inventor
Friedrich Emanuel Hust
Original Assignee
Mercedes-Benz Group AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mercedes-Benz Group AG filed Critical Mercedes-Benz Group AG
Publication of WO2023198393A1 publication Critical patent/WO2023198393A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/68—Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
    • H04N23/682—Vibration or motion blur correction
    • H04N23/685—Vibration or motion blur correction performed by mechanical compensation
    • H04N23/687—Vibration or motion blur correction performed by mechanical compensation by shifting the lens or sensor position
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/50—Monitoring users, programs or devices to maintain the integrity of platforms, e.g. of processors, firmware or operating systems
    • G06F21/55—Detecting local intrusion or implementing counter-measures
    • G06F21/554—Detecting local intrusion or implementing counter-measures involving event detection and direct action
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/70—Protecting specific internal or peripheral components, in which the protection of a component leads to protection of the entire computer

Definitions

  • the invention relates to a method for operating a vehicle.
  • image stabilization is known from the prior art, as described in https://de.wikipedia.org/wiki/Bildstabilmaschine (accessed on March 28, 2022). In photography, image stabilization is a process used to avoid camera shake.
  • the invention is based on the object of specifying a method for operating a vehicle that is improved compared to the prior art.
  • the object is achieved according to the invention by a method for operating a vehicle with the features of claim 1.
  • the vehicle has at least one camera with, in particular optical, in particular optomechanical, image stabilization.
  • the camera is one, for example Mono camera or stereo camera.
  • the image stabilization uses at least one acceleration sensor to detect camera fluctuations and initiates image stabilization measures based on this. In particular, the image stabilization compensates for motion blurring caused by the detected fluctuations by moving a sensor chip and/or a lens of the camera.
  • the camera's sensor chip is an image sensor, meaning it is used to capture images. It is therefore in particular provided that the camera has an, in particular optical, in particular optomechanical, image stabilizer with at least one such acceleration sensor and at least one device for displacing the sensor chip and/or the lens.
  • acceleration sensors can also be provided, for example three acceleration sensors, with each of these three acceleration sensors detecting fluctuations or portions of the fluctuations of the camera in the direction of one of the axes of a three-dimensional coordinate system.
  • a 4D transformation of signals from the acceleration sensors as a so-called quantum ion is also conceivable.
  • this camera is used to operate the vehicle, in particular to carry out a ferry operation of the vehicle, in particular to carry out an automated, in particular partially automated, highly automated or autonomous, ferry operation of the vehicle.
  • the camera is used in particular to detect the external surroundings of the vehicle, in particular to detect objects in the surroundings.
  • acoustic signals are detected which are coupled to the at least one acceleration sensor.
  • a detected acoustic signal is identified as an acoustic interference signal if it has a spectral component whose frequency corresponds to a natural frequency of the at least one acceleration sensor or a harmonic of this natural frequency or is in a predetermined range around this natural frequency or the harmonic of this natural frequency of the at least one acceleration sensor and if this spectral component preferably has an amplitude that exceeds a predetermined amplitude limit value.
  • the harmonic of the natural frequency is a frequency that is an integer multiple of the natural frequency.
  • the specified area or the specified The deviation limit value and the amplitude limit value are advantageously determined as part of test measurements in which it is determined up to which frequencies and amplitudes of the detected acoustic signal there is a noticeable disturbance in the image stabilization.
  • the predetermined deviation limit value of the deviation from the natural frequency can, for example, be on the order of approximately 10 kHz. In particular it is 4 kHz. If an acoustic interference signal is detected, ie an acoustic signal identified as an interference signal, it is concluded that the image stabilization is malfunctioning.
  • the method according to the invention thus enables malfunctions of the vehicle's camera equipped with image stabilization to be detected.
  • the malfunction can, for example, be based on an active acoustic attack on the vehicle, which can disrupt the operation of the vehicle. This can result, for example, in danger for the vehicle, especially for its occupants, and for other road users.
  • the solution according to the invention enables active detection and, for example, prevention of such an acoustic attack. For example, at least part of an attack vector of the interference signal can be mitigated. This can, for example, ensure safe automated, in particular autonomous, ferry operation of the vehicle.
  • a warning is issued to a user of the vehicle. This means that the user is warned and can, for example, independently initiate appropriate protective measures if necessary, for example monitoring the vehicle's automated ferry operation and, if necessary, intervene and take over driving the vehicle themselves.
  • an influence of the identified acoustic interference signal on the at least one acceleration sensor is determined and this influence is compensated for in a sensor signal of the at least one acceleration sensor. This means that negative effects of the acoustic interference signal on image stabilization can be avoided or at least reduced.
  • the vehicle is transferred to a safe state and/or the Image stabilization disabled.
  • the vehicle control of the vehicle is handed over to the user of the vehicle and/or a speed of the vehicle is reduced and/or the vehicle is steered to the edge of the road and /or the vehicle is stopped and/or one of the vehicle's hazard lights is activated.
  • the acoustic signals are detected by means of a microphone of a hands-free system and/or a voice control of the vehicle and/or by means of a microphone already installed in the vehicle for another purpose. In this embodiment, it is not necessary to install an additional microphone in the vehicle to carry out the method, which saves corresponding costs and corresponding assembly effort.
  • the acoustic signals are detected using a microphone that is installed in the vehicle exclusively for this purpose. A position of the microphone used for the method is in particular as close as possible to a location where the acoustic interference signal is impressed, i.e. H.
  • the microphone used for the method is arranged in particular in an interior, in particular in a passenger compartment, of the vehicle.
  • a communication connection is established to a device external to the vehicle, after a malfunction of the image stabilization has been concluded or when the vehicle is brought into a safe state.
  • This can be done, for example, by the so-called eCall function or in another way and can be, for example, a voice and/or data transmission connection.
  • the vehicle-external device can, for example, be staffed by human personnel and/or be a security event management system. In this way, for example, the current situation can be clarified and/or the user of the vehicle can be given information about the current situation and/or how to proceed.
  • the environmental data is recorded in particular by means of at least one environmental detection sensor or several environmental detection sensors of the vehicle, in particular before and/or during and/or after a malfunction of the image stabilization has been concluded.
  • This makes a later evaluation possible, in particular in order to be able to understand the situation that has occurred. For example, this makes it possible to determine whether it was a targeted active attack or a randomly occurring interference signal. If the interference signal occurs permanently at the position in which the vehicle was affected, precautionary measures can, for example, be taken when carrying out the automated ferry operation for the vehicle and, for example, also for other vehicles and/or the source of the interference signal can be determined so that an attempt can be made to switch off or shield the interference signal.
  • Show: 1 shows schematically the effect of an acoustic interference signal on a camera of a vehicle
  • FIG. 2 shows schematically an approach for detecting and compensating for an acoustic interference signal acting on at least one acceleration sensor of a camera
  • FIG. 3 shows schematically a detection of an acoustic signal using a microphone of a hands-free system
  • Fig. 6 schematically shows another example of determining the acoustic model
  • Fig. 7 shows schematically an output of a warning in the event of a malfunction of an image stabilization of the camera.
  • a method for operating a vehicle is described below with reference to FIGS. 1 to 7, for which the vehicle has at least one camera 1 with, in particular optical, in particular optomechanical, image stabilization 6.
  • this image stabilization 6 is combined with an image stabilization which, as mentioned at the beginning, is based on image processing supported by Kl.
  • the camera 1 is, for example, a mono camera or stereo camera.
  • the image stabilization 6 detects fluctuations SA of the camera 1 by means of at least one acceleration sensor 2 or several acceleration sensors 2 and compensates for the detected fluctuations SA by moving a sensor chip 3 and/or a lens of the camera 1.
  • three acceleration sensors 2 are provided, each of these three acceleration sensors 2 fluctuations SA or shares of the Fluctuations SA of the camera 1 are detected in the direction of one of the axes of a three-dimensional coordinate system.
  • this camera 1 is used to operate the vehicle, in particular to carry out a ferry operation of the vehicle, in particular to carry out an automated, in particular partially automated, highly automated or autonomous, ferry operation of the vehicle.
  • the camera 1 is used in particular to detect the external surroundings of the vehicle, in particular to detect objects O in the surroundings.
  • the problem with such image stabilizations 6 is that due to an acoustic interference signal SS acting on the respective acceleration sensor 2, the respective acceleration sensor 2 can malfunction and thus the image stabilization 6 can malfunction, in particular if a frequency of the acoustic interference signal SS is in the range a natural frequency of the respective acceleration sensor 2 lies.
  • 1 shows schematically an effect of such an acoustic interference signal SS on the acceleration sensor 2 of the camera 1.
  • the acceleration sensor 2 is arranged together with the sensor chip 3 for the image capture of the camera 1 on a circuit board 4 in a housing of the camera 1.
  • Light L for example, is incident on the sensor chip 3 via a lens of the camera 1.
  • the acoustic interference signal SS acts on the camera 1, whereby vibrations V are generated, which are detected by the acceleration sensor 2.
  • the acoustic interference signal SS can be, for example, a randomly occurring acoustic signal S or an acoustic interference signal SS of a targeted attack. Such an attack is also known as a poltergeist attack. Excitation of the respective acceleration sensor 2 by means of the acoustic interference signal SS can result, for example, in non-detection, incorrect detection or incorrect classification of objects O in the area surrounding the vehicle.
  • the solution described below enables, in particular, active detection and prevention of the consequences of such acoustic interference signals SS. For example, this makes it possible to mitigate at least part of a vector of the acoustic interference signal SS and thus ensure safer operation of the vehicle in automated ferry operations. This can be done in particular by merging the acceleration sensors 2 and the acoustic signal detected by a microphone 5 Interference signal SS can be achieved in order to compensate for any interference impressions in the respective acceleration sensor 2.
  • the basis of the problem caused by the acoustic interference signal SS is the excitation of the acceleration sensors 2, through which the stabilization of an image capture by means of the camera 1 is actually to be achieved and thereby object recognition OE and in particular object classification are to be made possible.
  • the acoustic interference signal SS is recognized. If an acoustic interference signal SS is detected, it is in particular provided that interference frequencies in a sensor signal of the respective acceleration sensor 2 are removed or at least reduced.
  • acoustic signals S are detected, which are coupled to the at least one or respective acceleration sensor 2.
  • a detected acoustic signal S is identified as an acoustic interference signal SS if its frequency corresponds to a natural frequency of the at least one or respective acceleration sensor 2 or is in the range of the natural frequency of the at least one or respective acceleration sensor 2.
  • the acoustic interference signal SS is detected, i.e. H. an acoustic signal S identified as interference signal SS, it is concluded that the image stabilization 6 is malfunctioning.
  • FIG. 2 shows the procedure described schematically using an exemplary embodiment.
  • the camera 1 is shown with the sensor chip 3, the acceleration sensors 2 and the image stabilization 6.
  • a traffic light is detected as object O in the area surrounding the vehicle using the sensor chip 3 of the camera 1.
  • the fluctuations SA of the camera 1 are detected and the detected fluctuations SA are compensated for by moving the sensor chip 3 and/or a lens of the camera 1, so that the capture of high-quality, in particular unshakable, images UB is made possible by means of the sensor chip 3 becomes.
  • These high-quality, especially unblurred, images UB are shown in the Example of object recognition OE and object classification, so that as a result E, the object O in the area surrounding the vehicle is recognized as a traffic light.
  • an acoustic signal S also acts on the camera 1, which is coupled into the acceleration sensors 2.
  • This acoustic signal S is detected and, in the example shown, passed into an acoustic model M, also referred to as an acoustic model, and into a threshold value detection SE, also referred to as threshold detection.
  • the acoustic model M calculates an acceleration caused by the acoustic signal S, which is detected by the respective acceleration sensor 2 and thus disturbs it. This acceleration is then compensated for in image stabilization 6, also known as image stabilization. An influence of the detected acoustic signal S on the respective acceleration sensor 2 is thus determined and this influence is compensated for in the sensor signal of the respective acceleration sensor 2.
  • the capture of high-quality, in particular unshakable, images UB of the detected object O is made possible by means of the sensor chip 3 of the camera 1.
  • this determination of the influence of the detected acoustic signal S on the respective acceleration sensor 2 and the compensation of this influence in the sensor signal of the respective acceleration sensor 2 takes place regardless of whether the detected acoustic signal S is identified as an acoustic interference signal SS or not. This also makes it possible to compensate for smaller impacts on the acceleration sensors 2, which are caused by acoustic signals S that are not identified as acoustic interference signals SS.
  • the threshold value detection SE is used to check whether the detected acoustic signal S has a spectral component whose frequency is close to the natural frequency of the respective acceleration sensor 2 or close to a harmonic of this natural frequency and whose amplitude is a predetermined amplitude limit value exceeds.
  • a frequency is close to the natural frequency or close to a harmonic of the natural frequency if it is in a predetermined range around the natural frequency or around the harmonic of the natural frequency.
  • the frequency is in particular close to the natural frequency or close to a harmonic of the natural frequency if its deviation from the natural frequency or the harmonic of the natural frequency is less than a predetermined deviation limit value.
  • the threshold value detection SE is used to check whether the frequency of the detected acoustic signal S is close to the natural frequency or a harmonic of the natural frequency of the respective acceleration sensor 2 and whether the amplitude of the detected acoustic signal S associated with this frequency exceeds the amplitude limit value exceeds. If this is the case, the detected acoustic signal S is identified as an acoustic interference signal SS and a malfunction of the image stabilization 6 is concluded. In this case, a warning W is issued to a user of the vehicle.
  • the influence of the detected acoustic signal S on the respective acceleration sensor 2 is determined and this influence is compensated for in the sensor signal of the respective acceleration sensor 2.
  • this only takes place after a malfunction of the image stabilization 6 has been concluded, i.e. H. upon detection of the acoustic interference signal SS, i.e. only after the detected acoustic signal S has been identified as an interference signal SS.
  • a microphone 5 can be used to detect the acoustic signal S, as shown in FIGS. 3 and 4.
  • the microphone 5, which is used for this is a microphone MF of a hands-free system 7 of the vehicle.
  • the microphone 5, MF is used via a low pass TP for the hands-free system 7 and via a high pass HP for the method described here, ie for detecting the acoustic signal S for the acoustic model M and the threshold value detection SE.
  • the microphone 5, MF here is therefore a dual-use microphone, with acoustic signals S being divided into low-frequency components and high-frequency components.
  • the acoustic signals S are thus divided by means of the low pass TP into a range suitable for human hearing, which can then be used, for example, for the hands-free system 7, and by means of the high pass HP into the range relevant for the interference with the acceleration sensors 2 and for that Methods described here are used, in particular for image stabilization 6 using of the acoustic model M during the action of the acoustic interference signal SS on the acceleration sensors 2 and as a detection mechanism for identifying the acoustic interference signal SS and triggering the warning W by means of the threshold value detection SE.
  • FIG. 4 shows an alternative embodiment in which two microphones 5, MF, in particular with local separation, for these two purposes, i.e. H. for the hands-free system 7 and for the method described here.
  • a microphone MF is provided for the hands-free system 7 and an additional microphone 5 is for the method described here, i.e. H. for detecting the acoustic signal S for the acoustic model M and the threshold value detection SE.
  • the position of the microphone 5 for the method described here i.e. H. for detecting the acoustic signal S, in particular the interference signal SS, is advantageously arranged as close as possible to an imprinting location, i.e. H. as close as possible to the camera 1, in particular to its respective acceleration sensor 2, thus for example in an upper area of a windshield in an interior, in particular passenger interior, of the vehicle, if the camera 1, designed for example as a mono camera or stereo camera, is arranged there. If this is not possible using the microphone MF already installed in the vehicle for the hands-free system 7, the embodiment according to FIG. 4 with the additional microphone 5 is particularly advantageous.
  • the detected acoustic signal S is identified as an acoustic interference signal SS in particular if its frequency is close to the natural frequency or a harmonic of the natural frequency of the respective acceleration sensor 2 and if the amplitude associated with this frequency is above the amplitude limit value, because then the disruptive effect on the respective acceleration sensor 2 is particularly large.
  • Characteristic oscillations then arise around the X-axis, Y-axis and/or Z-axis of the three-dimensional coordinate system in the vicinity of the natural frequency.
  • the acceleration Xcß(t), YGß(t), Zcß(t) measured by the acceleration sensors 2 can be decomposed into a real acceleration XRß(t), YRß(t), ZRß(t) and one caused by the acoustic interference signal SS Acoustically caused acceleration XAß(t), YAß(t), ZAß(t), each over time t:
  • X GB (t) X RB (t) + X AB (t) (1)
  • the acoustically caused acceleration and the measured acceleration are always coupled by the frequency impressed by the acoustic interference signal SS.
  • Y GB (t) Y RB (t) + Y AB (t) - f y (t)
  • Z GB (t) Z RB (t) + Z AB (t) - f z (t)
  • the acoustic model M can be stored as a characteristic curve.
  • acoustic stimuli i.e. H. Signals S
  • the acoustic signal S can be, for example, an acoustic sweep, i.e. H. a periodic signal with increasing frequency, or a single frequency excitation or an acoustic pulse can be used.
  • Other forms of excitation for example multi-frequency excitation, square, triangle, sawtooth or others, are also possible.
  • the triggered disturbance of the acceleration sensors 2 can then be described for each individual frequency, recorded in a lookup table 8, also known as a conversion table, and the acoustic model M can be described as a transfer function represent f x (t), f y (t), f z (t).
  • a camera measurement KM using a reference image RB, for example from ISO-16505, during the acoustic excitation by the acoustic signal S using the camera 1 and to use a deviation from the reference image RB to determine the relationship between the reference image RB, currently measured image and the acceleration, as shown in Figure 6.
  • a reference image RB for example from ISO-16505
  • the acoustic excitation of the camera 1 takes place by means of the acoustic signal S, whereby it is not accelerations that are measured using the acceleration sensors 2, but rather the effect of the acoustic excitation by the acoustic signal S on the image captured by the camera 1, that is, its deviation from the image Reference image RB due to the acoustic excitation by the acoustic signal S and the resulting disturbed image stabilization 6.
  • the acoustic model M can be represented as a transfer function f x (t), f y (t), f z (t).
  • the threshold detection SE it makes sense to issue a warning W, as shown in Figure 7.
  • the warning W can, for example, be acoustic and/or visual and can also be logged, for example.
  • the predetermined limit is in particular the range of the natural frequency of the respective acceleration sensor 2.
  • a detected acoustic signal S is thus identified as an acoustic interference signal SS if its frequency corresponds to the natural frequency of the respective acceleration sensor 2 or is in its range, whereby Detection of an acoustic interference signal SS indicates a malfunction of the image stabilization 6 and then the warning W is issued to the user of the vehicle. If, in the case of an acoustic interference signal SS, the values of the acoustic model M, which are required to compensate for the acoustic interference signal SS, are outside a tolerable range, it makes sense to transfer the vehicle to a safe state and advantageously also to deactivate the image stabilization 6. i.e.
  • the vehicle is transferred to a safe state and/or the image stabilization 6 is deactivated.
  • a situation-related assessment can be carried out. For example, it can be provided that the vehicle control of the vehicle is handed over to the user of the vehicle and/or a speed of the vehicle is reduced and/or the vehicle is steered to the edge of the road and/or the vehicle is stopped and/or a hazard warning light system of the vehicle is activated.
  • a communication connection is then established to a device external to the vehicle, for example the mentioned eCall function or in another way.
  • the vehicle-external device can, for example, be staffed by human personnel and/or be a security event management system, i.e. H. the aforementioned security incident management system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Computer Security & Cryptography (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Fittings On The Vehicle Exterior For Carrying Loads, And Devices For Holding Or Mounting Articles (AREA)

Abstract

Selon l'invention, - le véhicule comporte au moins une caméra (1) comprenant un moyen de stabilisation d'image (6) qui détecte un tremblement (SA) de la caméra (1) au moyen d'au moins un capteur d'accélération (2) et déclenche des mesures de stabilisation d'image sur la base de cette détection, - des signaux acoustiques (S) qui sont injectés dans le ou les capteurs d'accélération (2) sont détectés, - un signal acoustique détecté (S) est identifié en tant que signal d'interférence acoustique (SS) lorsqu'il a une composante spectrale dont la fréquence correspond à une fréquence naturelle du ou des capteurs d'accélération (2) ou à une harmonique de cette fréquence naturelle ou se trouve dans une région prédéfinie autour de cette fréquence naturelle ou autour de l'harmonique de la fréquence naturelle, et - un dysfonctionnement du moyen de stabilisation d'image (6) est déduit lorsqu'un signal d'interférence acoustique (SS) est détecté.
PCT/EP2023/056645 2022-04-12 2023-03-15 Procédé pour commander un véhicule WO2023198393A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102022001241.3A DE102022001241A1 (de) 2022-04-12 2022-04-12 Verfahren zum Betrieb eines Fahrzeugs
DE102022001241.3 2022-04-12

Publications (1)

Publication Number Publication Date
WO2023198393A1 true WO2023198393A1 (fr) 2023-10-19

Family

ID=85727172

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2023/056645 WO2023198393A1 (fr) 2022-04-12 2023-03-15 Procédé pour commander un véhicule

Country Status (2)

Country Link
DE (1) DE102022001241A1 (fr)
WO (1) WO2023198393A1 (fr)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102014221682A1 (de) * 2014-10-24 2016-04-28 Robert Bosch Gmbh Verfahren und Vorrichtung zum Betreiben eines Fahrzeugs
DE102019201689A1 (de) * 2019-02-11 2020-08-13 Zf Friedrichshafen Ag Verfahren und Steuereinheit zum Betreiben eines autonomen Fahrzeugs

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102012217002A1 (de) 2012-09-21 2014-03-27 Robert Bosch Gmbh Verfahren und Vorrichtung zum Betreiben eines Kraftfahrzeugs in einem automatisierten Fahrbetrieb
DE102013013867A1 (de) 2013-08-20 2015-03-12 Audi Ag Kraftfahrzeug und Verfahren zur Steuerung eines Kraftfahrzeugs
US9858621B1 (en) 2014-05-20 2018-01-02 State Farm Mutual Automobile Insurance Company Autonomous vehicle technology effectiveness determination for insurance pricing
WO2019207653A1 (fr) 2018-04-24 2019-10-31 三菱電機株式会社 Dispositif, procédé et programme de détection d'attaque
CN112333402B (zh) 2020-10-20 2021-10-22 浙江大学 一种基于声波的图像对抗样本生成方法及系统

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102014221682A1 (de) * 2014-10-24 2016-04-28 Robert Bosch Gmbh Verfahren und Vorrichtung zum Betreiben eines Fahrzeugs
DE102019201689A1 (de) * 2019-02-11 2020-08-13 Zf Friedrichshafen Ag Verfahren und Steuereinheit zum Betreiben eines autonomen Fahrzeugs

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
JI XIAOYU ET AL: "Poltergeist: Acoustic Adversarial Machine Learning against Cameras and Computer Vision", 2021 IEEE SYMPOSIUM ON SECURITY AND PRIVACY (SP), IEEE, 24 May 2021 (2021-05-24), pages 160 - 175, XP033964744, DOI: 10.1109/SP40001.2021.00091 *
TRIPPEL TIMOTHY ET AL: "WALNUT: Waging Doubt on the Integrity of MEMS Accelerometers with Acoustic Injection Attacks", 2017 IEEE EUROPEAN SYMPOSIUM ON SECURITY AND PRIVACY (EUROS&P), IEEE, 26 April 2017 (2017-04-26), pages 3 - 18, XP033113275, DOI: 10.1109/EUROSP.2017.42 *
Z. SHIF. SHIW. -S. LAIC. -K. LIANGY. LIANG: "Deep Online Fused Video Stabilization", 2022 IEEE/CVF WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV, 2022, pages 865 - 873

Also Published As

Publication number Publication date
DE102022001241A1 (de) 2023-10-12

Similar Documents

Publication Publication Date Title
DE102012222303B4 (de) Vorrichtung und Verfahren für das Entfernen von reflektiertem Licht von einem Bild eines Bildgebungsgeräts
DE102016215291A1 (de) Verfahren zur Klassifikation von Fahrerbewegungen
DE19749372A1 (de) Elektronisches Erkennungssystem und -verfahren für akustische Signale
EP1462820B1 (fr) Indication visuelle dans un système de collecte de données pour véhicule
DE102006028176A1 (de) Relativgeschwindigkeits-Rechenvorrichtung und Fahrzeug-zu-Fahrzeug-Abstands-Regelvorrichtung
WO2017102150A1 (fr) Procédé d'évaluation d'une situation de danger détectée par au moins un capteur d'un véhicule, procédé de commande d'une repoduction d'un avertissement de danger et procédé de reproduction d'un avertissement de danger
EP1988389A1 (fr) Surveillance d'une zone avec la détermination de la contamination d'une surface transparente basée sur le contraste d'image
DE102010024415B4 (de) Verfahren und Vorrichtung zur Aufnahme einer Bilderfolge einer Umgebung eines Fahrzeugs
DE102018222825A1 (de) Verfahren und System zum Betrieb eines Fahrzeugs mit akustischer Erfassung einer Fahrzeugaußenumgebung
WO2012017334A2 (fr) Système de caméra destiné à être utilisé dans des véhicules
DE102010049216A1 (de) Verfahren zum Betrieb einer an einem Fahrzeug angeordneten Kamera
EP3677019B1 (fr) Procédé et dispositif de commande prévisible d'exposition d'au moins une première caméra de véhicule
DE102015211443A1 (de) Verfahren und Vorrichtung zum Bereitstellen eines Bezugsniveaus für eine Augenöffnungsweite
DE102019116058A1 (de) Verfahren und vorrichtung zur objekterkennung in kamera-blindzonen
WO2023198393A1 (fr) Procédé pour commander un véhicule
DE102005026188A1 (de) Steuerverfahren für eine Unfallschutzeinrichtung in einem Fahrzeug
DE10318111A1 (de) Verfahren und Vorrichtung zur Erkennung eines Fahrzustands
DE102015006571A1 (de) Vorrichtung und Verfahren zur Ermittlung einer Verwindung einer Karosserie und/oder eines Fahrgestells
DE102011079005B4 (de) Verfahren zur Eigendiagnose von Kameras und Bildaufnahmesystem mit mindestens einer Kamera zur Durchführung des Verfahrens
DE102019204209A1 (de) Verfahren und System zum Bestimmen eines Funktionsstatus eines Mikrofons
DE102015009184A1 (de) Vorrichtung und Verfahren zur Vermeidung von Kollisionen eines Fahrzeugs mit einem Objekt
DE102009027231A1 (de) Vorrichtung zur Ortung von Objekten im Umfeld eines Fahrzeuges, insbesondere eines Kraftfahrzeuges, und Verfahren zum Betrieb einer derartigen Vorrichtung
DE102019213746A1 (de) Verfahren zur Steuerung des Betriebs eines Anbaugerätes
DE102014018364A1 (de) Vorrichtung und Verfahren zur Unterstützung eines Fahrers eines Fahrzeugs, insbesondere eines Nutzfahrzeugs
EP3493171A1 (fr) Détection du comportement agressif dans les moyens de transport publics

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23712830

Country of ref document: EP

Kind code of ref document: A1