Method and device for environmental detection

Abstract

Method for sensing the environment of a vehicle (2) comprising the steps: Operating at least two measuring devices (6, 7) which operate according to the pulse-echo measuring principle and have spatially overlapping measuring ranges, Fusion of the measurement results of the at least two measuring devices (6, 7) for object location in the vicinity (3) of the vehicle (2); where the at least two measuring devices (6, 7) perform their pulse-echo measurements simultaneously or quasi-simultaneously, so that in the recorded echo profile (31, 32) of a respective measuring device (6, 7) in addition to direct echo pulses (61, 62; 65, 66), which are caused by a reflection of the emitted measuring pulse (42, 43) of the respective measuring device (6, 7) on an object (4, 5) in the environment (3), also cross-echo pulses (63, 64; 67, 68) occur, which are caused by a reflection of a measuring pulse (42, 43) emitted by another measuring device (6, 7) on an object (4, 5) in the overlap area of ​​the measuring devices (6, 7); and a classification step is performed before merging in order to classify at least individual echo pulses (33) with regard to their membership and / or non-membership in a transmission path class (K) ij) to classify, and subsequently the fusion of the measurement results to locate objects (4, 5) is carried out taking into account the determined classification, wherein pulse-echo measurements are carried out quasi-simultaneously when the different at least two measuring devices (6, 7) emit their measurement pulses (42, 43) at different times, the time offset, however, is significantly shorter than the measurement duration for which echo pulses reflected after the emission of the measurement pulse (42, 43) are subsequently recorded with time resolution, characterized in that for the classification of cross-echo pulses (63, 64; 67, 68) the transmission paths between two measuring devices (M i ) a difference image space is determined by calculating, for each location (x, y) or each spatial region, the magnitude of the difference (D) of the signal strengths (Si(T) ij (x,y))) of the echo profiles (31, 33) of the two measuring devices (M i) to the signal propagation times (T) assigned to the location (x, y) or spatial area ij (x,y)) is determined, and in the resulting difference image space (9) elliptical traces (151-153) are determined, which are formed from pixels whose pixel intensity values ​​are above a threshold value, and for these elliptical traces (151-153) in the echo profiles (101-104) the associated echo pulses (107) are determined, which generated the corresponding elliptical traces (151-153), and these echo pulses (107) are removed from the echo profiles (101-104) for the measurement result fusion in connection with the transmission paths between these measuring devices (M i ) be used .

Description

[0001] The invention relates to a method and a device for detecting the environment of a vehicle using measuring devices which operate according to the pulse-echo measuring principle.

[0002] It is known from the prior art to implement environmental sensing for vehicles, such as motor vehicles, in order to locate or determine objects and / or open spaces in the vehicle's vicinity. This is important for motor vehicles, for example, in connection with parking maneuvers and parking assistance systems, collision warning systems, and similar functions.

[0003] Measuring devices that operate on the pulse-echo principle utilize a signal transducer that converts an electrical signal into an emitted measurement signal. This transducer can be, for example, an ultrasonic transducer that converts an electrical signal into an ultrasonic signal. Such an emitted measurement signal, in particular a measurement pulse, is radiated into the vicinity of the measuring device. If an object is located within the measuring range of the device, it reflects at least part of the incident measurement pulse back to the device. Typically, the transducer used to generate the measurement pulse is also used to detect these reflected echo pulses and convert them into an electrical signal. The measurement result of a pulse-echo measurement is available in the form of a time-resolved sampled echo signal, which is referred to as an echo profile.

[0004] From EP 2 090 897 A1, a measuring system is known which comprises a plurality of measuring devices that measure the area surrounding a vehicle using a pulse-echo measurement method. In addition to direct echoes, which are echo pulses detected by a measuring device and caused by a measurement pulse emitted by the same measuring device, the measuring system described therein also evaluates so-called cross-echoes or cross-pulse echoes. Cross-pulse echoes are echoes detected by one measuring device but caused by a measurement pulse emitted by another measuring device. The evaluation of the direct echo pulses and cross-echo pulses makes it possible to determine whether an obstacle is located directly in front of a measuring device or between two measuring devices.

[0005] A method and a device for collision prevention are also known from DE 10 2004 020 426 A1. Several sensor devices operating on the pulse-echo measurement principle are arranged on a pivotally mounted vehicle door. These monitor the surroundings to prevent the vehicle door from opening beyond a maximum predetermined opening angle or from colliding with an obstacle in the vicinity. Both direct echo pulses and cross-echo pulses can be evaluated to improve environmental detection.

[0006] WO 03 / 070517 A1 discloses a method for parking a vehicle in which both the distances of the vehicle to obstacles and the length and / or width of a parking space are determined. Measuring devices operating on the pulse-echo principle are used for both parking space determination and distance measurement. To validate measurements from one measuring device, the distance values ​​determined by another measuring device are used. To enable more precise localization of obstacles, cross-echo pulses can also be evaluated in addition to direct echo pulses.

[0007] From DE 10 2009 054 634 A1, a device for monitoring the distance between a vehicle and an object is known. The device described here is equipped with several measuring devices connected to an electronic unit in the front or rear of the vehicle. This measuring device also evaluates both direct echo pulses and cross-echo pulses. Objects in the vicinity are located by superimposing the pulse-echo measurement results.

[0008] It is known from the prior art to evaluate direct echo pulses and cross-echo pulses, each of which is correlated with a signal propagation time of an emitted measurement pulse. This propagation speed of the measurement pulse in the environment provides an indication of the distance traveled from the transmitting measuring device to the reflecting object and back to the receiving measuring device. Depending on the known geometry of the arrangement of the measuring devices relative to each other, objects in the vicinity of a vehicle can thus be located.

[0009] DE198 56 974C1 describes a device for detecting the presence of an obstacle near an object, comprising at least one transmitter located on the object for emitting a signal pulse, at least one pair of receivers located on the object for capturing one of the reflected echoes of the signal pulse, and an evaluation circuit for measuring the travel times of the captured echoes and for determining the presence of an obstacle in a first ambient area if the travel time to the first receiver is less than a third threshold value chosen to be greater than the first threshold value, and simultaneously the travel time to the second receiver is greater than a fourth threshold value chosen to be greater than the second threshold value.

[0010] DE 102004020423 A1 describes a method for detecting fictitious obstacles during distance measurements of a motor vehicle based on reflections, comprising the following steps: Determining echo band ranges, each assigned to an echo signal of a sensor pair consisting of two sensor devices, within which the distance values ​​of any fictitious obstacles assigned to the echo signals can occur; Determining limit values, each assigned to an echo band range of the individual echo signals of the sensor pair; Emitting transmission signals of the measuring medium by the two sensor devices of the sensor pair; Receiving the reflected echo signals of the measuring medium by the two sensor devices of the sensor pair; Checking for the occurrence of a first echo signal of the echo signals of the sensor pair within the echo band range assigned to the first echo signal;Checking for the occurrence of at least one of the other echo signals of the sensor pair below the respective assigned threshold values ​​of the echo band ranges assigned to the other echo signals, in the event that the first echo signal occurs within the band range assigned to the first echo signal; and detecting a fictitious obstacle in the event that the other echo signals of the sensor pair do not occur below the respective assigned threshold values ​​of the echo band ranges assigned to the other echo signals. In this method, the sensors are operated sequentially.

[0011] DE 101 24 909 A1 describes a method and a device for operating a radar sensor arrangement with a number of adjacent, largely synchronized individual sensors for determining the position of a target object. In a measurement cycle, the travel time of the radar signal emitted by an individual sensor and reflected by the target object to that individual sensor (direct echo) and to another individual sensor (cross echo) is evaluated. From the evaluation of the direct and cross echoes, at least the position of the target object is determined, a detection quality signal is calculated, and after a predetermined value of the detection quality signal is reached for a target object, a calibration is performed.

[0012] A method for evaluating ultrasonic sensors was also described by Fernando Moita et al. in “Mapping with a binaural system” in Proceedings 2003 IEE international Conference on Industrial Technology, 10-12 Dec. 2003 in Maribor, Slovenia, Vol. 1, pp. 117-122, ISBN 0-7803-7852-0.

[0013] A disadvantage of many known devices and methods from the prior art is that only one of several measuring devices can emit a measurement pulse at a time in order to unambiguously assign direct pulse echoes and cross-pulse echoes. If two or more measuring devices are operated in such a way that both direct pulse echoes and cross-pulse echoes occur in a recorded echo profile from one measuring device, measurement evaluation and data fusion become more difficult, and the unambiguousness of the localization results is eliminated or reduced.

[0014] German patent DE 10 2008 002 232 A1 discloses a method and a device for determining the distance and / or speed of an object relative to a vehicle. This method provides that a new measurement process of a pulse-echo measurement is initiated as soon as an echo pulse is received in the measuring device that emitted a measurement pulse or in another adjacent measuring device. This reduces the recording time for the individual echo profiles in situations where an object is located close to the measuring device. Echo profile data corresponding to signal propagation times associated with greater object distances generally do not need to be recorded in such cases, or are not to be expected, since the object often prevents the measurement pulse from propagating to these more distant areas within the measuring device's range.Although the described method increases the possible repetition rate of pulse-echo measurements, data acquisition is still limited.

[0015] The invention is therefore based on the technical problem of improving known measurement methods in such a way that a larger number of distance measurements can be recorded and meaningfully evaluated within a given time interval.

[0016] The invention is solved by a method for sensing the environment of a vehicle with the features of claim 1 and a device with the features of claim 7. Advantageous embodiments are described in the dependent claims.

[0017] The invention is based on the idea of ​​operating several measuring devices, i.e., at least two measuring devices operating according to the pulse-echo measurement principle, simultaneously or quasi-simultaneously, so that in addition to direct echo pulses, cross-echo pulses also appear in the echo profiles when objects are located in an overlapping area of ​​the measurement ranges of the at least two measuring devices. In order to utilize the additional information contained in the increased number of received echo pulses in such a way as to improve the localization result compared to similar measurements with prior art measuring systems, it is provided to classify at least individual echo pulses by determining their membership and / or non-membership in one or more transmission path classes and to take this classification into account when fusing the measurement results for locating objects in the environment.

[0018] In particular, a measurement method is proposed which comprises the following steps: operating at least two measuring devices that operate according to the pulse-echo measurement principle and have spatially overlapping measurement ranges; fusing the measurement results of the at least two measuring devices for object localization in the vicinity of the vehicle; wherein it is provided that the at least two measuring devices perform their pulse-echo measurements simultaneously or quasi-simultaneously, so that in the recorded echo profiles of the respective measuring device, in addition to direct echo pulses, which are caused by the reflection of the emitted measurement pulse of the respective measuring device on an object in the vicinity, cross-echo pulses also occur, which are caused by a reflection of a measurement pulse emitted by another measuring device on the one object in the overlap area of ​​the measuring devices;and a classification step is performed before fusion to classify at least individual echo pulses with regard to their membership and / or non-membership in a transmission path class, and subsequently the fusion of the measurement results to locate objects is carried out taking into account the determined classification.

[0019] If a fusion method is used in which individual probabilities for the presence of an object are determined for each location or spatial area in the environment based on the different transmission path classes, a classification of the echo pulses and their assignment to transmission paths between two measuring devices can be determined in such a way that, according to the invention, a difference image space is determined in which, for each location or spatial area, the magnitude of the difference in the signal strengths of the echo profiles of the two measuring devices at that time position corresponding to the signal propagation time assigned to the location or spatial area is determined, and elliptical traces are determined in the resulting difference image space, which are formed from pixels whose pixel intensity values ​​are above a threshold value, and the corresponding echo pulses are determined for these elliptical traces in the echo profiles.which generated the corresponding elliptical traces, and these echo pulses are removed from the echo profiles used for measurement result fusion in connection with the transmission paths between these measuring devices. While calculating this difference image represents an additional computational step, it allows, in a very simple and reliable way, the identification of those echo pulses in a pair of echo profiles that cannot be assigned to cross-echo pulses caused by the pair of measuring devices currently being considered as a pair.

[0020] For every echo pulse captured in an echo profile of a measuring device, at least one transmission path exists. A transmission path originates at a measuring device that emits the measurement pulse, leads to an object in the environment where the measurement signal is reflected, and from there to a measuring device that captures the reflected measurement pulse as an echo pulse. Depending on the object's position in the vehicle's vicinity, a multitude of transmission paths exist. Transmission paths where the transmitting measuring device and the measuring device receiving the echo profile are identical are grouped into a transmission path class. A transmission path class thus encompasses all transmission paths that lead to direct pulse echoes from a specific measuring device.Another transmission path class encompasses all transmission paths in which the measurement pulse is emitted by a measuring device, reflected by an object in the environment, and detected as an echo pulse in a specific other measuring device. The echo profiles appearing in the measurement result, i.e., the echo profile of the specific other measuring device, for this transmission path class are all cross-echo pulses. A unique identification of a transmission path class is possible, for example, by indexing the measuring devices and assigning an index pair to each transmission path class, where the first index indicates the transmitting measuring device and the second index indicates the receiving measuring device. i,i This indicates a transmission path class.

[0021] If the indices are different (i≠j), the transmission paths correspond to cross-echo pulses. If the indices are the same (i=j), the transmission paths correspond to direct echo pulses.

[0022] A corresponding device for environmental sensing comprises at least two measuring devices that operate according to a pulse-echo measuring principle, wherein the measuring devices are designed and arranged so that they have a spatially overlapping measuring range, and a control and evaluation device which is designed to control the at least two measuring devices and to evaluate their acquired measurement results in the form of echo profiles and to perform object localization, wherein it is provided that the control and evaluation device is designed to control the measuring devices in such a way that they perform their pulse-echo measurements simultaneously or quasi-simultaneously and the control and evaluation device is designed to classify at least individual acquired echo pulses with regard to an associated transmission path and to take the classification into account when fusing the measurement results for object localization.

[0023] The advantage of the invention is that measurement data fusion for object localization can be performed much more precisely if, at least for individual echo pulses in the recorded echo profiles, an assignment to the respective transmission path is possible, or the belonging to one of the transmission paths can be excluded.

[0024] When the measuring devices are operated in such a way that they perform their pulse-echo measurements simultaneously, the emitted measurement pulse is transmitted by at least two measuring devices at the same time. The emitted measurement pulses are generally identical in terms of intensity, frequency, and duration. Slight variations may occur due to variations in the transducers used in the measuring devices or due to influences from the immediate environment at the installation location of the measuring device in a vehicle.

[0025] Near-simultaneous pulse-echo measurements are defined as measurements in which the different measuring devices transmit their measurement signal, i.e., their measurement pulse, with a time delay. However, this time delay is significantly shorter than the measurement duration for which the reflected echo pulses are subsequently recorded with time resolution. For ultrasonic transducers, such as those commonly used in motor vehicles, typical delay times between consecutively transmitted measurement pulses from different measuring devices range from 0.5 to 5 ms, typically from 0.7 to 1.5 ms.

[0026] If multiple objects are present in the vicinity of the vehicle, several echo pulses belonging to the same transmission path class can naturally occur within a single echo profile. For example, two objects spaced apart and located at different distances from a measuring device, which do not shadow each other with respect to the rectilinear propagation of the measurement signal, each generate a direct echo pulse in the recorded echo profile. The time difference with which the direct echo pulses are recorded in the echo profile is correlated with the difference in distance between the two objects and the measuring device.

[0027] When operating several measuring devices simultaneously or quasi-simultaneously, the number of echo pulses recorded in the echo profiles naturally increases, although it is possible that no measurable echo pulses can be found in the echo profiles for individual theoretically possible transmission path classes, because, for example, the reflection properties of the object are unfavorable for certain reflection angles or the transmission path is "obstructed" by other objects.

[0028] In order to achieve the best possible object localization and to use as much of the information contained in the captured echo profiles as possible, one embodiment provides that, for different transmission path classes for generating a direct echo pulse or a cross-echo pulse, the signal propagation times are determined or provided for a set of locations or spatial areas in the vicinity of the vehicle, which correspond to the respective direct echo pulse or cross-echo pulse of the respective transmission path that is generated in the echo profile.If an object is present in the vicinity of the vehicle only at the corresponding location or in the corresponding spatial area, and if, during measurement data fusion, individual probabilities for the presence of an object are determined for one or more of the transmission path classes for the locations or spatial areas based on the signal intensity in the echo profiles associated with the transmission path classes, which occur at those times that correspond to the signal propagation times of the respective transmission path class assigned to the locations or spatial areas. For example, if the transmission path class considered is the one that combines those transmission paths that originate from one measuring device, lead to an object, and return to the same measuring device, then each location in the vicinity is assigned an individual probability value for the presence of an object at that location.in which this probability is derived from the signal intensity of the echo profile of this measuring device, whereby the intensity at the time in the echo profile is used that corresponds to the signal propagation time assigned to the corresponding location for the transmission path. If the echo profile shows no or low intensity, this indicates that there is no object at this position that triggers a reflection, or that the transmission path to this location is blocked by an object located closer to the measuring device. If the individual probabilities thus obtained for the different transmission path classes for a location are added together,This results in an overall probability for the presence of an object at this location. In total, this provides a probability statement about the presence of an object for every location in the vicinity. Graphically, a location result can be represented, for example, by encoding the probability of an object's presence at each location / area in the environment using a color value. This creates an intensity map of the environment. Areas with colors associated with a high probability indicate the presence of an object at that location / area.

[0029] Since the pulse-echo measurement principle only allows for the conclusion that, for example, in the case of a direct echo, an object is located on a circular arc with a radius corresponding to the measured signal propagation time, concrete object localization can only be achieved by fusing echo pulses from different transmission path classes. This is done by adding the individual probabilities for the different transmission path classes. Locations or areas where an object is located will, after such fusion, exhibit higher probability values ​​than those areas or locations for which an increased probability of an object's presence is indicated solely based on a detected echo pulse.However, if every echo pulse occurring in the echo profile is considered in this evaluation for each of the transmission path classes relevant to the echo profile, then incorrect probability values ​​for the presence of an object will inevitably be assigned to locations and spatial areas. If, however, at least some echo pulses in an echo profile are classified beforehand, these classified echo pulses are not considered in the evaluation of individual transmission path classes and / or are considered exclusively in other transmission classes. This results in a significant improvement in object localization with this evaluation method.

[0030] Since the length of a transmission path belonging to a cross-echo pulse from one measuring device to the object and to the other measuring device is the same as the transmission path from the other measuring device to the same object and to the first measuring device, there are two transmission path classes. If one measuring device and the other measuring device transmit their measurement pulse simultaneously, each of these classes will cause a cross-echo pulse at the same time in the two recorded echo profiles of the two measuring devices. To classify cross-echo pulses, one can therefore search the corresponding echo pulse profiles of the measuring devices associated with the transmission path classes for similar echo pulses that occur at the same time and with nearly identical signal strength. This again assumes that the transmitted measurement pulses are also similar with regard to intensity, duration, etc.

[0031] In one embodiment, the echo profiles of the measuring devices are examined in pairs for similar echo pulses with corresponding signal propagation times. Such similar echo pulses are classified as cross-echo pulses of the transmission path classes between the measuring devices of the corresponding pair, and only these echo pulses are used in the measurement result fusion with respect to the transmission path class between these two measuring devices. If the individual measuring devices are not operated simultaneously but quasi-simultaneously, the echo profiles must be corrected accordingly with respect to the start time differences of the measurements, i.e., with respect to the time difference in the emission of the measurement pulse.

[0032] In one embodiment, a search for similar echo pulses to locate cross-echo pulses is performed by adding the echo pulse profiles, corrected for the different start times of the measurement pulse transmission, in pairs to form sum profiles. Echo pulses whose intensities, within a specified tolerance threshold, correspond to half of the sum profile at the time position assigned to the respective echo pulse (where the respective echo pulse appears in the echo profile), are classified as cross-echo pulses of the transmission path classes between the two measuring devices from whose echo profiles the sum profile is formed. Preferably, only echo pulses above a predetermined threshold are considered in order to suppress random noise.

[0033] A nearly complete classification of all received echo profiles is possible with an embodiment in which the individual measuring devices are operated quasi-simultaneously. Here, the quasi-simultaneously performed pulse-echo measurements of the at least two measuring devices form a so-called measurement sequence. For environmental sensing, measurement sequences are repeatedly executed. Crucially, the order in which the individual measuring devices emit their measurement pulses in successive measurement sequences is varied. In a first measurement sequence, the measuring devices, which are indexed (e.g., numbered) for easier evaluation and explanation, are triggered to emit their measurement pulse at equidistant time intervals, for example, at intervals of 1 ms.In the subsequent measurement sequence, the measuring devices are then, for example, caused to emit their measurement pulses in reverse order, also at equidistant time intervals. However, the time interval between the emission of two measurement pulses in the second measurement sequence is preferably different from the time interval between two measurement pulse emission in the first measurement sequence. For example, the measurement pulses in the second measurement sequence are emitted at intervals of 1.25 ms. Assuming that the measuring device relative to the environment, and the environment itself, has not changed between the measurements performed in the different measurement sequences, the echo profiles of both measurement sequences contain essentially the same information.The term "essential" is chosen here because, in this specific case, depending on the geometric arrangement of the measuring devices and the objects relative to each other, as well as the chosen time intervals for transmitting the measurement signals in the echo profiles, echo signal superposition can occur, which can lead to a reduction in information. Regardless of these potentially occurring "complications," in principle, the echo profiles of the sequentially acquired measurement sequences of a measuring device contain "the same" echo pulses. "The same echo pulses" here means that echo pulses occur along the same transmission paths. Due to the delay in the measurement pulse transmissions, cross-echo pulses can be "lost," arriving at a measuring device before it begins acquiring echo pulses.

[0034] For evaluation purposes, it is advantageous to define a common absolute zero point for the different recorded echo profiles of a measurement sequence, which, for example, coincides with the start of the emission of the measurement pulse by a selected measuring device. This selected measuring device is used in all measurement sequences to define the zero point of the respective measurement sequence.

[0035] For example, if the first measurement pulse is emitted by the distinguished measuring device in the first measurement sequence, the echo profiles of the other measuring devices in the first measurement sequence must each be corrected by the amount corresponding to the time delay for the emission of the measurement pulse by the respective measuring device.

[0036] If, for example, another measuring device transmits its measurement pulse 1 ms after the designated measuring device in the first measurement sequence, then an echo pulse detected 10 ms after the transmission of this measurement pulse by the other measuring device is detected 11 ms after the time zero of the measurement sequence. If, for example, the other measuring devices transmit their measurement pulses before the designated measuring device in the second measurement sequence, their detected echo profiles must be corrected by a negative time difference.

[0037] Regarding the evaluation for determining direct echo pulses, it is only necessary to compare the echo profiles of the same measuring device from two consecutive measurement sequences, ensuring that the measurement pulse was emitted at the same time each time. In the echo profiles of the same measuring device, acquired in consecutive measurement sequences, the direct echo signals exhibit the same interval from the time of emission of the measurement pulse by the measuring device capturing the echo profile. Such echo pulses occurring at identical times can therefore be classified as direct echo pulses.

[0038] This is most easily achieved by binaryizing the acquired echo profiles with respect to a discrimination threshold. If the intensity of an acquired echo pulse is above the discriminator threshold, the echo pulse is replaced by a standard pulse. If the intensity of an echo pulse is below the discriminator threshold, the echo profile in the area of ​​the acquired echo signal is set to the value 0. Thus, the binary normalized echo profile only contains standard pulses, which, for example, are assigned the value 1 and are located at those points in the echo profile where a maximum with an intensity above the discriminator threshold is recorded. Echo pulses occurring at the same time positions in the successively acquired echo profiles of the same measuring device can be selected by multiplying the binaryized and / or normalized echo profiles together, i.e., by creating a product profile.Only if a standard echo pulse is present in both profiles at the same time position will an echo pulse also be detectable in the product profile. The pulses occurring in the product profile are therefore classified as direct echo pulses.

[0039] To determine the cross-echo profiles in an analogous manner, it is necessary to correct the recorded echo profiles with respect to the transmission times of the measurement pulses causing the cross-echo pulses in the respective measurement sequences. This means that the echo profiles must be corrected with respect to the transmission times of another measuring device that transmits this measurement pulse in each successive measurement sequence. After this correction, the cross-echo pulses for a given transmission path class are located at the same time position in the echo profiles of the single recording measuring device, which are recorded in the different measurement sequences. Locating these echo pulses at the same time positions is again easily possible by binaryizing the echo profiles before or after correcting the time offset, and if necessary, further normalizing them, thus creating a product echo profile.Remaining binary or, if applicable, normalized echo pulses identify cross-echo pulses of the considered transmission path class. To locate the different cross-echo pulses, the corresponding transmission path classes or pairs of measuring devices and their echo profiles must be evaluated. Subsequently, at least all echo pulses whose intensity is above the discriminator threshold for binaryization are assigned to a transmission path. In the subsequent measurement result fusion, only the corresponding echo pulses assigned to each transmission path under consideration are used. Alternatively, echo pulses that do not belong to a particular transmission path can be removed from the echo profile used for that transmission path during fusion.

[0040] In the device according to the invention, the control and evaluation unit is designed to carry out a method as described above in conjunction with the at least two measuring devices.

[0041] In one embodiment, a memory is provided in which the signal propagation times assigned to the individual locations or spatial areas of the environment for the different transmission path classes are or were stored in tabular form. For each transmission path class, there is a table in which the assigned signal propagation time for the different locations or spatial areas is stored. The transmission paths from one measuring device to another have the same signal propagation times as the transmission paths from the other measuring device to the first measuring device.Thus, for the cross-echo pulses between two measuring devices, although these belong to different transmission path classes, only one transmission path table is necessary, whereby in connection with the evaluation it is necessary to correct the signal propagation times with regard to the possibly existing time offset due to the different measurement pulse emissions of the different measuring devices.

[0042] The invention is explained in more detail below with reference to embodiments and a drawing. The drawing shows: Fig. 1 a schematic representation of a device for environmental sensing; Fig. 2 a schematic representation of two signal propagation delay tables for two different transmission path classes; Fig. 3 the recorded echo profiles from four measuring devices; Fig. 4 the corresponding standardized echo profiles for the four measuring devices; Fig. 5 a graphical representation of a location result of objects in the environment; Fig. 6 graphical representations of the object probabilities for individual transmission path classes; Fig. 7 a graphical representation of a difference image space to check whether echo pulses occurring in echo profiles can be assigned to selected transmission path classes; Fig. 8 a schematic representation of a localization result taking into account the classification of individual echo pulses; and Fig. 9a - 9c a schematic representation of received echo profiles in successive measurement sequences as well as schematic evaluations for classifying direct echo pulses ( Fig. 9b) a measuring device and for classifying cross-echo pulses ( Fig. 9c) the same measuring device with respect to another measuring device.

[0043] In Fig. Figure 1 schematically shows a device 1 for environmental sensing, preferably in a vehicle 2. The environment 3 of the vehicle 2 is examined for possible objects 4, 5. The device 1 is designed to locate these objects 4, 5 in the environment 3 of the vehicle 2.

[0044] The environmental sensing device 1 comprises at least two measuring devices 6, 7, which operate according to the pulse-echo measuring principle. Preferably, the measuring devices are ultrasonic measuring sensors that emit an ultrasonic measuring pulse by means of an ultrasonic transducer (not shown) and receive echo pulses generated by reflection from the objects 4, 5 and convert them into electrical signals. These electrical signals are recorded with time resolution and are referred to as echo profiles.

[0045] The measuring devices 6, 7 are coupled to a control and evaluation unit 8, which controls the measuring devices 6, 7 and acquires and evaluates the measurement results acquired by the measuring devices 6, 7. The control and evaluation unit 8 typically comprises a microprocessor 11 and a memory unit 12 containing the program code, which is executable on the microprocessor 11 and defines the functionality of the control and evaluation unit 8. To locate the objects 4, 5 in the environment 3, various evaluation steps are performed on the measurement results from the measuring devices 6, 7. These steps can be executed, for example, by means of the program-controlled microprocessor 11. Individual processing steps for the measurement results can also be performed in a field-programmable gate array or a hard-wired special circuit to accelerate signal processing.In some embodiments, the microprocessor can also be completely replaced by a special circuit for signal evaluation. The functionality of the environmental sensing device and the evaluation and control unit is explained in more detail below. Location results are provided via an interface 13. This can be, in particular, a bus interface, but also any other interface suitable for exchanging data. The interface can also be a software interface if the control and evaluation unit is implemented using hardware that provides further functionalities.

[0046] At the in Fig. In the situation depicted in Figure 1, the device 1 for environmental sensing comprises two measuring devices 6, 7, and two objects 4, 5 are located in the environment 3. The following section briefly explains the schematic appearance of the acquired measurement results in the form of echo profiles 31, 32, which are acquired by the two measuring devices 6, 7. The echo pulses shown are for illustrative purposes only. The time intervals shown do not correspond exactly to those in Figure 1. Fig. The geometry shown in Figure 1. However, echo profiles 31 and 32 are based on the structure shown in Figure 1. Fig. The geometry shown in Figure 1 is described. Furthermore, the origin of the echo pulses 33 appearing in the schematically depicted echo profiles 31, 32, which are sometimes also referred to as echo pulses, and their assignment to different transmission paths are explained.

[0047] The echo profile 31 shown above depicts the time-resolved signal recorded at the signal converter of the measuring device 6, while the echo profile 32 shown below depicts the time-resolved signal recorded at the signal converter of the measuring device 7. The measuring devices 6 and 7 are controlled by the control and evaluation unit 8 such that they simultaneously emit a measurement pulse 42, 43 into the environment 3. In the example shown, the measurement pulse is emitted as an ultrasonic signal, which propagates radially from the measuring devices 6 and 7 in a circular sector. These circular sectors (not shown) overlap and approximately define the measuring range of each measuring device 6 and 7. Echo profile 31 schematically shows the intensity of the emitted and received pulses plotted against the time axis. At time t0 41, the measurement pulse 42 is emitted by the measuring device 6.The echo pulses 33 detectable in the echo profile 31 of the measuring device 6 originate from different transmission paths. The measuring pulse 42 propagates radially from the measuring device 6.

[0048] A transmission path 51 leads from the measuring device 6 to the object 4 and back to the measuring device 6. The echo generated by this transmission path 51 is a so-called direct echo and produces a direct echo pulse 61. A transmission path 52 begins at the measuring device 6, leads to the object 5 and back to the measuring device 6, and generates a direct echo pulse 62 in the echo profile 31. Another transmission path 53, which is partially identical to transmission path 51, leads from the measuring device 6 to the object 4 and then to the measuring device 7. At the measuring device 7, the reflected echo pulse thus generates a so-called cross-echo pulse 63 in the echo profile 32 of the measuring device 7. Similarly, a transmission path 54 exists, which leads from the measuring device 6 to the object 5 and then to the measuring device 7, resulting in the formation of the cross-echo pulse 64 in the echo profile 32 of the measuring device 7.Similarly, starting from the measurement pulse 43, which is emitted by the measuring device 7 at time t0 41, there exist transmission paths 55 to 58, which accordingly lead to the direct echo pulses 65, 66 through reflections on the objects 4 and 5 in the echo profile 32 of the measuring device 7 as well as to cross-echo pulses 67, 68 in the echo profile 31 of the measuring device 6 due to the reflections on the objects 4, 5.

[0049] The different transmission paths can be assigned to different transmission path classes. For example, transmission paths 51 and 52, which lead to direct echo pulses 61 and 62 in the echo profile 31 of the measuring device 6, can be assigned to the transmission path class for direct echo pulses for the measuring device 6. Additionally, in other embodiments with more than two measuring devices, there is a further transmission path class for each additional measuring device whose measurement pulse can generate cross-echo pulses. Transmission paths 57 and 58, which lead to the cross-echo pulses 67 and 68 in the echo profile 31 of the measuring device 6, belong to such a transmission path class for cross-echo pulses. One possible notation for the different transmission path classes K is to identify them by two indices i and j. ij, where the first index i indicates the sending measuring device and the second index j indicates the receiving measuring device. If measuring device 6 is assigned index 1 and measuring device 7 index 2, these can accordingly also be designated as measuring devices M1 and M2. Thus, for example, the transmission path class K includes 22 the transmission paths 55, 56 of the direct echo pulses 65, 66 of the measuring device 7 (M2). The transmission path class K 12 includes the transmission paths 53, 54, which generate the cross-echo pulses 63, 64 in the echo profile 32 of the measuring device 7 (M2).

[0050] It is easy to see in Fig. 1. The cross-echo pulses 63, 64 in the echo profile 32 of measuring device 7 (M2) occur at the same times t2, t3, 44, 45 as the cross-echo pulses 67, 68 in the echo profile 31 of measuring device 6 (M1). This property exists when the two measuring devices 6, 7 emit their measurement pulses 42, 43 simultaneously and can be used to classify the cross-echo pulses. This will be explained in more detail below.

[0051] To deduce the positions of objects 4 and 5 in the environment 3 from the measurement results, which are available in the form of echo profiles 31 and 32, a so-called measurement result fusion is performed. One method is based on a holographic approach for object localization through trilateration. For each transmission path class, the signal propagation time is determined for the locations / spatial areas in the environment associated with a transmission path of that transmission path class whose reflection point is located at that location / in that spatial area. Typically, the environment is discretized for this purpose, i.e., a grid of points is overlaid on the environment to calculate only a limited number of points. For these points, which can also be considered, for example, as the centers of spatial areas that are, for instance, square in shape, the signal propagation times for the transmission path class can be stored in a table.The signal propagation time (dimensionless in sampling steps) is calculated according to the following formula: tj,i(x,y)=(x−Mx,i)2+(y−My,i)2+(x−Mx,j)2+(y−My,j)2c⋅Fs

[0052] Here, F s the sampling frequency at which the converter signal is sampled in the measuring devices, c is the speed of sound and M xi , M yi , M xj , M yj are the position coordinates (x, y) of the measuring devices (i, j). Examples are shown in Fig. 2. The determined signal propagation times are graphically represented. The signal propagation time belonging to a location is indicated by a hatching density. The higher the hatching density, the shorter the signal propagation time belonging to a transmission path whose reflection point is the corresponding location in the graphic representation. On the left side of the Fig. Figure 2 shows the determined signal propagation times for the direct echoes of a first measuring device M1 implemented in the form of an ultrasonic transducer, and on the right-hand side the corresponding signal propagation times for the transmission path class that leads to cross echoes of class K. 21 This is caused by a measurement pulse emitted by measuring device M2 in the echo profile of measuring device M1. Since the cross-echo pulses arrive at both measuring devices simultaneously when the measurement signals are emitted at the same time, the table corresponding to transmission path class K is relevant. 12 belongs, identical. The graphically represented tables are set up for an embodiment in which four measuring devices are arranged equidistant from each other along a straight line.

[0053] In Fig. Figure 3 shows a graphical representation of the echo profiles 101 to 104, which were acquired by the four measuring devices M1 to M4 designed as ultrasonic transducers. While echo profiles 101 and 102 of measuring devices M1 and M2 each contain very intense echo signals 105 with high intensity, echo profiles 103 and 104 contain only signals 106 with low intensity. It has proven advantageous to search specifically for echo pulses in the individual echo profiles. In one embodiment, this is done by means of a maxima search using a threshold value. Only maxima that lie above a threshold value are considered relevant echo pulses 107. In this way, the individual echo profiles 101 to 104 are, for example, binaryized in one embodiment and then convolved with a standard pulse.This offers the advantage that all detected echo pulses 107 are weighted equally during measurement data fusion, regardless of the actual measured echo pulse intensity. Nevertheless, even with this method, measurement data fusion is performed depending on the intensity occurring at a specific position in an echo profile, since the standard pulses depend on the measured intensity. Other embodiments can omit the normalization step and perform the evaluation with the original, unnormalized echo profiles, as described below.

[0054] In Fig. Figure 4 shows the standardized echo profiles 111 to 114 as described above, which are shown in conjunction with echo profiles 101 to 104 of the Fig. 3 correspond. Based on the tables for the signal propagation times of the individual transmission path classes, a measurement data fusion is now performed. Using the table, for example, for the direct echo pulse transmission path of the measuring device M1 (measuring device M1-object-measuring device M1), as shown in Fig. As shown in Figure 2, each position is assigned a value based on the signal propagation time contained in the table. This value depends on the intensity in the echo profile or normalized echo profile of the measuring device M1 at the time corresponding to the signal propagation time. For example, locations assigned a signal propagation time from the table for which no intensity is assigned in the normalized echo profile at that time are assigned the value zero. Conversely, locations whose signal propagation times coincide with time positions at which one of the normalized echo pulses occurs are assigned values ​​that correspond to the intensity at the respective positions in the normalized echo profile 111. Thus, for each transmission path class, a diagram is created in which the intensity values ​​over the area indicate a measure of the probability of an object's presence based on an echo profile.Adding these individual probability images together yields a graphical representation, as shown in . Fig. Figure 5 shows a multitude of circular and elliptical traces 121, and particularly high intensities arise at intersection points 122, characterized by a high hatching density. The higher the intensity, the greater the probability that an object is located in the vicinity at that point.

[0055] However, since cross-echoes occur alongside direct echoes in the individual echo profiles (101-104; 111-114), and these cross-echoes can even be caused by several different measuring devices, it inevitably follows that "false traces" occur in the described evaluation methods, in which the individual recorded echo pulses are not assigned to a transmission path and thus classified. Therefore, according to the invention, it is provided that at least some recorded echo pulses are classified and that the echo pulses thus classified are not taken into account when evaluating transmission path classes to which they do not belong, i.e., removed from the respective echo profile, and / or that they are used exclusively when evaluating the transmission path class to which they are assigned via the classification.

[0056] The simplest method of classification involves searching for echo pulses in the recorded echo profiles that are captured at the same time and produce approximately the same signal intensity. This approach assumes that the transmission paths between the measuring devices via reflection from an object (i.e., those that cause cross-echo pulses) are identical not only in terms of path length but also in terms of reflection intensity. It is further assumed that the measurement pulses are emitted simultaneously by the different measuring devices and exhibit identical or similar pulse width and pulse intensity.

[0057] Another way to identify cross-echo pulses of a transmission path is to consider a difference image space for each corresponding transmission path class between a pair of measuring devices. This is described as an example for the cross-echo pulses that may occur between the first measuring device M1 and the third measuring device M3, i.e., transmission path class K. 13 and K 31 In Fig. Figure 6 shows the two image spaces assigned to these channels, which represent the individual probabilities for a cross-echo pulse of transmission path class K. 31 and transmission path class K 13 to represent. Since the cross-echo impulses in the two echo profiles 101, 103 and the normalized echo profiles 111, 113 would have to occur at the same time, similar elliptical structures 141–143 would have to appear in both image spaces. This is, as already shown from Fig. As can be seen in Figure 6, this is not the case here. To perform this comparison of the image spaces, it is advantageous to calculate a difference image space for each pair of cross-correlations, i.e., pairs of measuring devices. The calculation is performed according to the following equation: D(x,y)=|Si(Ti,j(x,y))−Sj(Tj,i(x,y))|

[0058] S here represents the signal of the normalized echo profile or the echo profile of the respective measuring device Mi. The corresponding image space of such a difference is shown by way of example in Fig. Figure 7 shows three elliptical traces 151-153 with high intensity. The echo pulses 161-163 (compare Fig. 3) which caused these traces, i.e., led to a high difference intensity, therefore cannot be cross-echo pulses between the two measuring devices M1 and M3 under consideration. The normalized echo pulses 171-173 in the normalized echo profiles 111 and 113 can thus be classified as not being cross-echo pulses of class K. 13 and K 31It should be noted here that classification, as described here, also includes excluding individual echo pulses from specific transmission path classes, which already leads to a significant improvement in localization results. If this is carried out for all possible cross-echo configurations, the echo pulses can be classified by process of elimination to such an extent that only those echo pulses actually attributable to the respective cross-echo transmission paths are considered in the evaluation. Those echo pulses that show no trace or only a trace of low intensity in the difference image space during pairwise evaluation can also be clearly assigned to a cross-echo transmission path. The remaining, unclassified echo pulses, i.e., those that cannot be assigned to any cross-echo transmission path, can necessarily be assigned to the respective direct echo pulse transmission paths.

[0059] In Fig. Figure 8 shows a location result that takes the determined classifications into account. A significantly improved location result is obtained, as can be seen from the comparison of the Fig. 5 and Fig. 8 clearly results.

[0060] As an alternative to the classification method described above, cross-echo pulses can also be identified by first creating a summed echo profile from the echo profiles involved in the transmission paths whose cross-echo pulses are to be classified. This is again done in pairs for each individual measuring device. Those echo pulses within the echo profiles that exhibit an intensity equal to half the intensity in the summed profile at the corresponding position are classified as cross-echo pulses. Furthermore, it is advantageous to also require a minimum echo pulse intensity to discriminate against interfering signals occurring in the echo profiles. This ensures that only "true" echo pulses are classified.

[0061] The procedure described so far has always assumed that the measurement pulses emitted by the various measuring devices are transmitted simultaneously. If this is not the case, but rather they are transmitted with a short time offset that is shorter than the expected signal propagation times for receiving echo pulses, this time offset must be taken into account during the evaluation. This can be done, for example, by adjusting the signal propagation time tables for the different transmission paths accordingly. In this case, the cross-pulse echo transmission paths, or...If signal propagation timestamps are no longer identical, for example, if the measurement signal from measuring device M2 is transmitted later than the measurement signal from measuring device M1, then the cross-echo pulses in the echo profile of measuring device M1 appear later in the echo profile (relative to the transmission time of the measurement pulse from measuring device M1) than the corresponding cross-echo pulse in the echo profile of measuring device M2 (again, relative to the transmission time of the second measurement pulse from measuring device M2). Such a time-shifted transmission of the measurement pulses can also be used in other ways to classify the echo pulses.

[0062] The overlapping measurements of the measuring devices are referred to as a measurement sequence. Measurement sequences are repeatedly executed in succession. In these sequences, the time intervals and / or the order of the measurement pulse transmissions from the various measuring devices are varied. For example, in the first measurement sequence, the measuring devices are triggered to transmit their measurement pulses at staggered intervals of 1 ms, according to their indexing. In the subsequent measurement sequence, the order is reversed, and the time interval between measurements is increased to 1.25 ms. The second measurement sequence begins with the transmission of the measurement pulse from measuring device M4. This is followed by transmissions from measuring devices M3, M2, and M1, respectively, after 1.25 ms, 2.5 ms, and 3.75 ms.

[0063] For explanatory purposes, the following notation is used. Time points in the time base of one of the measuring devices are marked with a "^". The start of the measurement pulse transmission of the corresponding measuring device coincides with the zero point of the corresponding time base. At time t̂ = 0, the measuring device transmits its measurement pulse.

[0064] Time values ​​t without superscripts indicate time relative to a time base linked to the measurement sequence. A zero point of this time base coincides, for example, with the emission of a measurement pulse by one of the measuring devices in the measurement sequence. The measurement times at which the individual measuring devices begin emitting their measurement pulses are specified relative to this time base. Here, it is assumed that a zero point of the "general" time base of a respective measurement sequence coincides with the first measurement pulse emission in that sequence, regardless of which measuring device emits the pulse. The emission times of the measurement pulses are then specified in the following notation: t i,l where i represents the measuring device and l the measurement sequence. For the two measurement sequences mentioned above, the following applies: first measurement sequence :t 1,1 = 0 ms, t 2,1=1 ms, t 3,1 =2 ms, t 4,1 =3 ms, and for the second measurement sequence: t 4,2 =0 ms, t 3,2 =1.25 ms, t 2,2 =2.5 ms, t 1,2 =3.75 ms.

[0065] The echo profiles are each determined by a function of time S i,l specified, where i represents the measuring device and l the measuring sequence.

[0066] In Fig. Figure 9a schematically depicts echo profiles 201-1 and 202-1 from two measuring devices i and j in a first measurement sequence l=1, and echo profiles 201-2 and 202-2 from a second, subsequent measurement sequence l=2. For the sake of simplicity, these are already binary. The notation "-1" and "-2" appended to reference symbols indicates membership in a first or second measurement sequence, respectively. The echo profile 201-1 corresponds to the function S. i,1 and with the Echo profile 202-1 the function S j,1The echo profile 201-2 corresponds to the function S i,2 and with the Echo profile 202-2 the function S j,2 .

[0067] The individual echo profiles are displayed according to their time offset relative to the general time base of the respective measurement sequence. They are thus presented as they would be captured in the general time base, for example, by a multi-channel acquisition unit of the control and evaluation system.

[0068] Both the general time base 205 of each measurement sequence and the time bases of the measuring devices 206-x, where x represents i or j, are indicated by arrows. The points where the arrows are drawn mark the zero point of the time scale. Furthermore, the zero points of the time bases of the measuring devices coincide with the emission of the corresponding measurement pulse 42-1, 42-2, 43-1, 43-2 of the measuring device Mi, Mj. For the sake of simplicity, the zero points of the general time bases 205-1, 205-2 are shown aligned with each other.

[0069] This means that the measurement pulses 42 and 43 plotted in the spectra, which are emitted, reflect the respective time offset in the respective measurement sequence. Furthermore, the echo profiles 201 to 202 are already shown in binary form, with the emitted measurement pulses 42 and 43 displayed with increased intensity to identify them as emitted pulses. These are, of course, not considered further in the classification of the echo pulses.

[0070] The classification of the echo pulses and echo profiles of the measuring device Mj will be explained as an example: To identify the direct echo pulses 211-1 and 211-2 in the echo profiles 202-1 and 202-2 of the measuring device Mj, these are superimposed in time so that the emissions of the measurement signals 43-1 and 43-2 coincide. This is in Fig. Figure 9b is shown as an example. The shifted echo profiles 202D-1 and 202D-2 are shown. If a product echo profile 231 is formed, the only remaining echo pulses are e jj the direct echo pulses 211 of transmission path class K jj To achieve temporal superposition of the echo profiles of the two successive measurement sequences, the echo profiles are adjusted relative to the time base of the measurement sequence with respect to the time delay of the measurement pulse emission of "their measurement pulse", relative to the zero point of the measurement sequence. corrected. The echo profiles S thus corrected d j,l 202D-1, 202D-2 in the time base of the measurement sequence are Sdj,l(t)=Sj,l(t−tj,l)

[0071] The product profile for direct echo pulse evaluation e d j (t) 231 is then given by edj(t)=Sdj,1(t)*Sdj,2(t)=Sj,1(t−tj,1)*Sj,2(t−tj,2).

[0072] A temporal position of the occurring cross-echo pulses 212-1, 212-2 of transmission path class K ij The time base of the measurement sequence depends solely on the respective transmission time t. i,l The measurement pulse 42-1, 42-2 of the measuring device Mi is dependent. The echo profiles 202-1, 202-2 are therefore dependent with respect to the transmission time t. i,l The measurement pulse 42-1, 42-2 of the measuring device Mi in the respective measurement sequence l is corrected. Essentially, all echo profiles of a measurement sequence are shifted together so that the time of the measurement pulse emission of the measurement sequence 201-1, 201-2 of the measuring device Mi coincides with the zero point of the general time base. If this is done in both measurement sequences, the cross-pulse echoes 212-1, 212-1 caused by the measuring device Mi appear at the same temporal position in the shifted echo profiles 202K-1, 202K-2.

[0073] The corrected echo profile functions S Kijj,l The results are therefore: SKijj,l(t)=Sj,l(t−tj,l)

[0074] A product profile for cross-echo analysis e kii j (t) is given by eKijj(t)=SKijj,1(t)*SKijj,2(t)=Sj,1(t−ti,1)*Sj,2(t−ti,2).

[0075] In Fig. Figure 9c shows the correspondingly correctly shifted echo profiles 202K-1 and 202K-2. A product echo profile 232 is generated again, in which in this case only the cross-echo pulses e i,j of transmission class K ij remain.

[0076] This evaluation can be performed for all echo profiles and possible transmission paths in order to classify all echo pulses. It should be noted that individual echo pulses may be superpositions of different echo pulses. This is naturally not apparent in the binarized echo profiles; however, individual echo pulses can thus be assigned to different transmission path classes and are taken into account accordingly during the measurement data fusion, which is subsequently carried out, for example, analogously to the process described above.

[0077] It will be apparent to those skilled in the art that only exemplary embodiments are described here to illustrate the invention, and that the different described methods can be combined to implement the invention advantageously. In particular, the different described methods for classifying individual echo pulses can be combined and carried out together, for example, to increase the plausibility of the classification. Reference symbol list 1 Device for environmental sensing 2 vehicles 3 Environment 4, 5 object 6, 7 Measuring device 8 Control and evaluation unit 11 Microprocessor 12 Storage device 31 Echo profile (from measuring device 6) 32 Echo profile (from measuring device 7) 33 echo pulses 41 Time t0 42, 43 Measurement pulse 44 Time t2 45 Time t3 51-58 Transmission paths 61, 62 Direct echo pulses 63, 64 cross-echo impulses 65, 66 direct echo pulses 67, 68 Cross-echo impulses 101-104 Echo profiles 105 echo signals 106 signals 107 detected echo impulses 111-114 normalized echo profiles 121 tracks 122 Intersection point 141 - 143 elliptical structures 151-153 elliptical traces 161-163 Echo impulses 171-173 normalized echo pulses 201-x Echo profile of a first measuring device Mi in the measurement sequence x 202-x Echo profile of the measuring device Mj in a measurement sequence x 202D echo profile corrected for direct echo pulse evaluation 202 Echo profile corrected for cross-echo pulse evaluation 205 general time base of a measurement sequence 206-x Time base of a measuring device x 210 echo pulses caused by measurement pulse (42-I) of the measuring device i 211 Direct echo pulse caused by measuring pulse (42-I) of the measuring device i 212 Cross-echo pulse caused by measuring pulse (42-1) of the measuring device i 220 echo pulses caused by measurement pulse (43-I) of the measuring device j 221 Direct echo pulse caused by measuring pulse (43-I) of the measuring device j 222 Cross-echo pulse caused by measuring pulse (43-1) of the measuring device j 231 Product echo profile (direct echo pulses) 232 Product echo profile (cross-echo pulses)

Claims

[1] Method for sensing the environment of a vehicle (2) comprising the steps: Operating at least two measuring devices (6, 7) which operate according to the pulse-echo measuring principle and have spatially overlapping measuring ranges, Fusion of the measurement results of the at least two measuring devices (6, 7) for object location in the vicinity (3) of the vehicle (2); where the at least two measuring devices (6, 7) perform their pulse-echo measurements simultaneously or quasi-simultaneously, so that in the recorded echo profile (31, 32) of a respective measuring device (6, 7) in addition to direct echo pulses (61, 62; 65, 66), which are caused by a reflection of the emitted measuring pulse (42, 43) of the respective measuring device (6, 7) on an object (4, 5) in the environment (3), also cross-echo pulses (63, 64; 67, 68) occur, which are caused by a reflection of a measuring pulse (42, 43) emitted by another measuring device (6, 7) on an object (4, 5) in the overlap area of ​​the measuring devices (6, 7); and a classification step is performed before merging in order to classify at least individual echo pulses (33) with regard to their membership and / or non-membership in a transmission path class (K) ij) to classify, and subsequently the fusion of the measurement results to locate objects (4, 5) is carried out taking into account the determined classification, whereby pulse-echo measurements are carried out quasi-simultaneously when the different at least two measuring devices (6, 7) emit their measurement pulse (42, 43) at different times, the time offset, however, is significantly shorter than the measurement duration for which echo pulses reflected after the emission of the measurement pulse (42, 43) are recorded with time resolution, characterized by , that for the classification of cross-echo pulses (63, 64; 67, 68) the transmission paths between two measuring devices (M i ) a difference image space is determined by calculating, for each location (x, y) or each spatial region, the magnitude of the difference (D) of the signal strengths (Si(T) ij (x,y))) of the echo profiles (31, 33) of the two measuring devices (M i) to the signal propagation times (T) assigned to the location (x, y) or spatial area ij (x,y)) is determined, and in the resulting difference image space (9) elliptical traces (151-153) are determined, which are formed from pixels whose pixel intensity values ​​are above a threshold value, and for these elliptical traces (151-153) in the echo profiles (101-104) the associated echo pulses (107) are determined, which generated the corresponding elliptical traces (151-153), and these echo pulses (107) are removed from the echo profiles (101-104) for the measurement result fusion in connection with the transmission paths between these measuring devices (M i ) be used . [2] Method according to claim 1, characterized by , that for different transmission path classes (K ij) to generate a direct echo pulse (61, 62; 65, 66) or to generate a cross-echo pulse (63, 64; 67, 68) each for a set of locations or spatial areas in the vicinity (3) of the vehicle the signal propagation times (Ti j (x,y)) are determined or provided, which correspond to the corresponding direct echo pulse (61, 62; 65, 66) or cross-echo pulse (63, 64; 67, 68) of the respective transmission path, which are generated in the echo profile (31, 32) if an object (4, 5) is present in the environment (3) of the vehicle (2) only at the corresponding location (x, y) or only in the corresponding spatial area, and during measurement data fusion for one or more of the transmission path classes (K ij) for the locations (x, y) or spatial areas, individual probabilities for the presence of an object (4, 5) are determined based on the signal intensity in the echo profiles (31, 32) that occur at those times corresponding to the signal travel times (Ti) assigned to the locations (x, y) or spatial areas. j (x,y)) of the corresponding transmission path class K ij correspond. [3] Method according to one of claims 1 or 2, characterized by , that the echo profiles (101-104) of the measuring devices (M i ) are examined in pairs for similar echo pulses with corresponding signal propagation times, and such similar echo pulses are considered cross-pulse echoes of the transmission paths between the measuring devices (M i ) of the corresponding pair are classified and only these echo pulses are used in the measurement result fusion with respect to the transmission paths between these two measuring devices (M i ) be used. [4] Method according to any one of claims 1 to 3, characterized by , that the echo profiles of the different measuring devices are corrected with respect to the different times of emission of the measurement pulses (42, 43) of the pulse-echo measurements. [5] Method according to one of claims 4, characterized by , that the echo profiles corrected with respect to the difference in the times of emission of the measurement pulses (42, 43) are added pairwise to sum profiles and echo pulses whose intensities correspond within a further tolerance threshold to half of the sum profile at the time position assigned to the respective echo pulse, at which the respective echo pulse occurs in the echo profile, are classified as cross-echo pulses of the transmission paths between the two measuring devices from whose echo profiles the sum profile is formed. [6] Method according to any one of claims 1 to 5, characterized by, that the quasi-simultaneous pulse-echo measurements of the at least two measuring devices (M i , M j ) form a measurement sequence (I) and are measured by at least two measuring devices (M i , M j ) repeated measurement sequences (I) are performed, wherein the quasi-simultaneous pulse-echo measurements of the at least two measuring devices (M) are performed i , M j ) in successive measurement sequences (I) in different order their respective measurement pulses (42, 43) at time points (t n , t jl) transmit, wherein the temporal zero points of the echo profiles in the measurement sequences are temporally rigidly correlated with the respective transmission time of a distinguished measuring device of the at least two measuring devices, and for classifying the echo pulses the recorded echo profiles are converted into a binary signal depending on the signal intensity relative to a threshold value and subsequently the binary echo profiles with respect to the time offset (t0- t i,l ) of the transmission time (t jl , t il ) of the transmit pulse are corrected to a zero point in the respective measurement sequence and multiplied together to determine the direct echo pulses, whereby the echo pulses occurring in the product profile are classified as direct echoes, and to determine the cross-echo pulses of a transmission path class K ij the binary echo profiles of the measuring device M3 with respect to the time difference (t0- t il ) of the transmission time (til ) the measuring device (i) which emitted the measuring pulse, which emitted the cross-echo pulse (e ij ) has been generated, corrected to the zero point of the measurement sequence, and the binary echo profiles thus corrected (SilKij,Si(l+1)Kij) successive measurement sequences (I, I+1) are multiplied together, and the resulting product echo profile (e ji (t)) occurring echo pulses as cross-echo pulses (e ij ) be classified. [7] Device (1) for environmental sensing comprising at least two measuring devices (6, 7) which detect echo profiles (31, 32) according to a pulse-echo measuring method, wherein the measuring devices (6, 7) are designed and arranged such that they have a spatially overlapping measuring range, and a control and evaluation unit (8) which is designed to control the at least two measuring devices (6, 7) and to evaluate their recorded measurement results in the form of echo profiles (31, 32) and to perform object localization, where the control and evaluation unit (8) is configured to control the measuring devices (6, 7) in such a way that they perform their pulse-echo measurements simultaneously or quasi-simultaneously, and the control and evaluation unit (8) is configured to classify at least individual detected echo pulses with regard to an associated transmission path and to take the classification into account when fusing the measurement results for object localization, wherein pulse-echo measurements are performed quasi-simultaneously if the different at least two measuring devices (6, 7) emit their measurement pulse (42, 43) with a time offset, but the time offset is significantly shorter than the measurement duration for which echo pulses reflected after the emission of the measurement pulse (42, 43) are detected with time resolution. characterized by, that the control and evaluation unit (8) is designed for the classification of cross-echo pulses (63, 64; 67, 68) of the transmission paths between two measuring devices (M i ) to determine a difference image space by calculating, for each location (x, y) or each spatial region, the magnitude of the difference (D) of the signal strengths (Si(T) ij (x,y))) of the echo profiles (31, 33) of the two measuring devices (M i ) to the signal propagation times (Ti) assigned to the location (x, y) or spatial area j(x,y)) is determined, and in the resulting difference image space (9) elliptical traces (151-153) are determined which are formed from pixels whose pixel intensity values ​​are above a threshold value, and for these elliptical traces (151-153) in the echo profiles (101-104) the associated echo pulses (107) which generated the corresponding elliptical traces (151-153) are determined, and these echo pulses (107) are removed from the echo profiles (101-104) which are used for the measurement result fusion in connection with the transmission paths between these measuring devices (M i ) be used. [8] Device (1) according to claim 7, characterized by , that the control and evaluation unit (8) is configured to perform a method according to claims 1 to 6 in conjunction with the at least two measuring devices (6, 7). [9] Device (1) according to claim 7 or 8, characterized by, that a storage device (12) is provided in which the signal propagation times assigned to the individual locations (x, y) or spatial areas of the environment (3) for the different transmission paths are stored in tabular form. [10] Device (1) according to any one of claims 7 to 9, characterized by , that the control and evaluation unit (8) has a program-controlled microprocessor (11) to perform the temporal correction of the echo profiles (31, 32) and / or a binaryization of the echo profiles and / or a classification of echo pulses and / or a determination of probabilities for the presence of an object (4, 5) at one of the locations (x, y) or in one of the spatial areas.

Images