JP6311211B2 - Obstacle detection device - Google Patents

Description

  The present invention relates to an obstacle detection device, and more particularly to an obstacle detection device that reduces clutter by filtering reflected waves received by a radar.

  Conventionally, various obstacle detection (detection) systems have been provided in vehicles in order to prevent occurrence of collision accidents with objects or obstacles (other vehicles, pedestrians, etc.) outside the vehicle. In such a system, for example, a radar device such as a millimeter wave radar that transmits and scans a radar radio wave to a monitoring area around the vehicle and receives a reflected wave thereof is mounted on the vehicle, and is received by the radar device. Based on the reflected wave of the radar radio wave, obstacles such as other vehicles and pedestrians existing in the area around the vehicle are detected (Patent Document 1).

JP2013-019684A

  Conventionally used millimeter wave radars can detect dangerous objects even at night or in bad weather when the visibility of a driver alone or the detection performance of a visible camera is reduced, and are effective in improving cognitive support performance. However, since the detection performance varies depending on the travel environment and the type of the target (target), erroneous detection may occur.

  It is an object of the present invention to provide an obstacle detection apparatus capable of reducing the fluctuation in detection performance according to the traveling environment and increasing the accuracy of target detection in an apparatus for detecting an obstacle using a millimeter wave radar. And

  The above problem is solved by the present invention having the following features. That is, the obstacle detection device as one aspect of the present invention generates a beat signal using a driving environment acquisition unit that acquires a driving environment related to the host vehicle and an FMCW millimeter wave radar, and is derived from the beat signal. The analysis means for generating the distance spectrum using the frequency spectrum and generating the received power distribution indicating the probability density for each received power based on the distance spectrum, and the filter for removing the received power distribution due to the traveling environment Filter means for determining according to the acquired traveling environment and filtering the generated received power distribution by the determined filter, and an obstacle for detecting an obstacle based on the filtered received power distribution And an obstacle detection device.

  According to the present invention configured as described above, in the obstacle detection executed based on the received power distribution generated using the FMCW millimeter wave radar, the traveling environment related to the own vehicle is acquired, and the acquired traveling environment is acquired. And a received power distribution is filtered by the determined filter. Thus, by performing filtering using a filter according to the traveling environment, clutter caused by the traveling environment can be reduced.

  As one aspect of the present invention, the analysis unit generates a received power distribution based on the generated plurality of the distance spectra.

  According to the present invention configured as described above, a reception power distribution in which temporal fluctuations are averaged is generated, so that a reception power distribution with reduced temporal fluctuations can be generated.

  As one aspect of the present invention, the analysis means generates a received power distribution based on the plurality of distance spectra generated each time the host vehicle moves a very small distance.

  According to the present invention configured as described above, a reception power distribution in which environmental factors such as multipath are changed without changing a theoretical reception power value is generated, and therefore, a plurality of distances generated at the same position Rather than using a spectrum, it is possible to generate a received power distribution that takes into account fluctuation characteristics that change from moment to moment in an actual in-vehicle running state.

  As one aspect of the present invention, the analysis means generates each received power distribution based on data within a plurality of predetermined distance ranges of the distance spectrum, and the filter means travels corresponding to each predetermined distance range. A filter for removing the received power distribution due to the environment is determined according to the acquired traveling environment, and the generated received power distribution is filtered by the determined filter.

  According to the present invention configured as described above, a signal is extracted for each predetermined distance range on the distance spectrum, and each predetermined distance is obtained from the ratio of the received power of the distance spectrum within the extracted predetermined distance range. A reception power distribution for each range is generated. Similarly, the filter is a filter generated for each of a plurality of predetermined distance ranges. Thus, by performing processing for each distance range from the host vehicle, the accuracy of obstacle detection can be further increased.

  As one aspect of the present invention, the filter is generated based on the received power distribution generated by the analysis unit when there is no obstacle for each traveling environment.

  According to the present invention configured as described above, a reception power distribution when no obstacle exists for each traveling environment is generated by the analysis unit, and the reception power distribution is removed based on the generated reception power distribution. Generate a filter to

  According to the present invention, it is possible to reduce the variation in detection performance according to the traveling environment and increase the accuracy of target detection.

It is a schematic block diagram of the obstruction detection apparatus by embodiment of this invention. It is a functional block diagram of the obstacle detection apparatus by embodiment of this invention. It is a figure explaining the time change of the transmission frequency of a FMCW system millimeter wave radar, and a receiving frequency. It is a figure explaining the time change of the beat frequency of the millimeter wave radar of a FMCW system. It is a figure which shows the frequency spectrum by embodiment of this invention. It is a figure which shows the distance spectrum which converted the frequency axis of the frequency spectrum of FIG. 4a into the distance axis. It is a figure which shows the distance spectrum which converted the frequency axis of the frequency spectrum of FIG. 4a into the distance axis. It is a figure which shows the received power distribution produced | generated from the ratio of the received power in the partial distance range of the some distance spectrum by embodiment of this invention. This is a received power distribution representing radar detection performance in consideration of variations in the received power intensity of the radar. It is a figure which shows the distance spectrum produced | generated when the own vehicle is located in the point A by embodiment of this invention. It is a figure which shows the distance spectrum produced | generated when the own vehicle is located in the point B which moved the micro distance from the point A by embodiment of this invention. It is a processing flow of the obstacle detection apparatus by embodiment of this invention.

  Hereinafter, an obstacle detection device according to an embodiment of the present invention will be described with reference to the accompanying drawings. In each drawing, for the sake of convenience, the aspect ratio and shape of data may be expressed differently from actual ones.

  FIG. 1 is a schematic configuration diagram of an obstacle detection apparatus 1 according to an embodiment of the present invention. As shown in FIG. 1, the obstacle detection apparatus 1 mainly includes a control unit 10, a radar 20, and a camera 30, and these components are electrically connected. The obstacle detection device 1 can share the above-described components with other devices or other systems.

  The control unit 10 includes a processing unit 11, a storage unit 12, and an interface unit 13. For example, the control part 11 is comprised by ECU (Electronic Control Unit) in a vehicle.

  The processing unit 11 includes a processor, for example, and performs various processes by executing a program stored in the storage unit 12. The storage unit 12 includes a nonvolatile memory such as a flash memory, and a volatile memory such as an SRAM and a DRAM. The interface unit 13 is interposed between the processing unit 11 and other devices, and relays data exchange between them.

  The radar 20 is, for example, a millimeter wave radar (frequency 76 GHz to 77 GHz), and includes a front antenna for measuring a front area of the vehicle (own vehicle). The radar 20 transmits a measurement wave from a front antenna and receives a reflected wave reflected by an object (obstacle) such as another vehicle, a fixed structure on the road, or a pedestrian object, It is configured to measure the relative distance and relative speed with the vehicle. The vehicle is not limited to a four-wheeled vehicle but includes a traveling body such as a two-wheeled vehicle or a bicycle. The front antenna may be a transmission / reception antenna, or a transmission antenna and a reception antenna may be provided. The radar 20 can further include a rear antenna.

  The camera 30 is an image sensor or the like, and can acquire image information around the vehicle by imaging the surroundings in front of the vehicle including the measurement range of the radar 20. For example, the camera 30 is an in-vehicle camera, and is set on the back side of a rearview mirror and the like, and by photographing the front through the front window, it is possible to photograph other vehicles, the environment of the running road, and the like.

  The obstacle detection apparatus 1 can further include an output unit (not shown). The output unit is electrically connected to the control unit 10 and can output information for notifying the driver when an obstacle is detected. For example, the output unit includes a speaker and a display, and when the car navigation system is mounted on the vehicle, the output unit may be shared with the system.

  The obstacle detection device 1 can further include a communication unit (not shown). The communication unit is configured to be able to communicate with a server (not shown) disposed in an information distribution center that distributes driving support information to each vehicle, and can transmit and receive various types of information to and from the server. For example, the communication unit may include a wireless antenna, and when the car navigation system is mounted on the vehicle, the communication unit may be shared with the system.

  FIG. 2 shows an example of a functional block diagram of the obstacle detection apparatus 1 according to the embodiment of the present invention. The obstacle detection device 1 includes a traveling environment acquisition unit 41, a storage unit 42, an analysis unit 43, a filter unit 44, and an obstacle detection unit 45. Since these functions are realized by the components shown in FIG. 1, the functional block classification is not limited to the above.

  The traveling environment acquisition unit 41 can acquire a traveling environment related to the host vehicle. For example, using the image information acquired by the camera 30, whether the traveling environment related to the host vehicle is a crossroad (four-way), whether it is a T-junction (three-way), whether there is a signal, the front of the vehicle By identifying whether there is a building, whether there is a parked vehicle, etc., information on the traveling environment is acquired. By using the camera 30, it is possible to identify not only information on intersections, signals, buildings, and the like that hardly change over time, but also information on parked vehicles and the like that frequently change over time.

  In the above embodiment, the camera 30 is used. However, the present invention is not limited to this. For example, for obtaining information about intersections, signals, buildings, etc. that hardly change over time, car navigation mounted on the vehicle is used. The system can be used. In the car navigation system, the position of the host vehicle on the map information can be recognized by the GPS, so that information on intersections, signals, buildings, etc. that hardly change over time can be acquired.

  The storage means 42 has a function of storing various data such as spectrum data, power distribution data, and filter data in the storage unit 12. When the obstacle detection apparatus 1 includes a communication unit, the storage unit 42 acquires and stores data from a server, an optical disk, or the like via the communication unit. The storage means 42 can store a filter database, which will be described later, in the storage unit 12.

  The analysis means 43 generates a beat signal from the transmission wave and reception wave signals using an FMCW millimeter wave radar.

  FIG. 3A is a diagram showing temporal changes in the transmission frequency and reception frequency of the FMCW millimeter-wave radar. The change in the transmission frequency of the transmission wave is indicated by a solid line, and the object located at a distance R from the vehicle (radar). The change in the reception frequency of the reflected wave reflected from is shown by a broken line. The horizontal axis t represents time, and the vertical axis f represents frequency. As shown in FIG. 3a, a modulation signal obtained by multiplying a continuous wave by a triangular wave and frequency-modulated is used as a transmission signal, the center frequency of the transmission signal is f0, the modulation frequency of the triangular wave is fm, and the modulation width of the triangular wave is ΔF. Yes.

  In the FMCW system, a beat signal is obtained by mixing a part of a transmission signal with respect to a reception signal. FIG. 3b is a diagram showing the time change of the beat frequency fb of the beat signal when the transmission signal and the reception signal of the millimeter wave radar change with time as shown in FIG. 3a, and t1 to t5 in FIG. 3a and FIG. Means the same time.

At this time, the relative distance R and the relative speed V between the host vehicle and the object use the above-described f0, fm, ΔF, light speed c, beat frequency f _{ub in} the up section, and beat frequency f _db in the down section. The following formulas 1 and 2 are known.
R = (f _db + f _ub ) · c / (8ΔF · fm) (Formula 1)
V = (f _db −f _ub ) · c / (4f0) (Formula 2)

From Equation 1, it is understood that the frequency (f _db + f _ub ) on the frequency spectrum of the beat frequency has a linear relationship with the distance R.

  The analysis unit 43 performs a complex FFT operation on the generated beat signal, thereby generating a frequency spectrum (power spectrum) having, for example, the received power as the vertical axis and the frequency as the horizontal axis. The frequency spectrum is generated every one cycle (1 / fm) of the triangular wave, that is, every one sweep. In this case, one sweep is one measurement cycle. FIG. 4a is a diagram showing a frequency spectrum in a certain measurement cycle. In this example, a peak can be confirmed around the frequency f1. However, it is also possible to generate one frequency spectrum based on data obtained by a plurality of sweeps (for example, five sweeps), that is, using a plurality of sweeps as one measurement cycle.

Based on Equation 1, the analysis unit 43 generates a distance spectrum by converting the frequency axis of the generated frequency spectrum into a distance R axis. FIG. 4b is a diagram showing a distance spectrum obtained by converting the frequency axis of FIG. 4a into a distance axis. f1 and r1 have the relationship of f _db + _fub and R in Equation 1. The analysis unit 43 generates a distance spectrum (see FIG. 4b) every time a frequency spectrum (see FIG. 4a) is generated. As a result, a plurality of distance spectra are generated as the measurement time elapses.

In general, the received power Pr of the reflected wave (received wave) from the object is expressed by the following equation 3 using the transmission power Pt, the antenna gain G, the wavelength λ, the radar cross section σ, and the distance R to the object. It can be obtained by the radar equation shown.
Pr = Pt × G ² × λ ² × σ / {(4π) ³ × R ⁴ } (Formula 3)

  If the received power in the above equation is higher than the reception sensitivity determined by the performance of the system, the object can be detected. In such a case, when an object is present at a position away from the host vehicle by the distance r1, a peak can be confirmed around the distance r1 on the distance spectrum as shown in FIG. 4b.

  The analysis unit 43 calculates the horizontal axis with the vertical axis as the probability density, for example, from the ratio of the received power in a part of the distance range (distance range on the distance R axis) of the plurality of distance spectra generated by the predetermined number of measurement cycles. A received power distribution indicating a probability density for each received power is generated as received power. Specifically, the received power (P1,..., PN) at each distance (for example, each distance in discrete data) within a partial distance range (X1 to X2) of each generated distance spectrum is extracted. The received power distribution is generated by expressing the extracted ratios (for example, −110 dB: 2%, −109 dB: 5%,...) On the received power axis. In addition, it is preferable that the distance range (X1-X2) on the R-axis extracted from each distance spectrum is the same distance range. By generating one received power distribution using a plurality of distance spectra in this way, a received power distribution that averages temporal fluctuations is generated. Therefore, a received power distribution with reduced temporal fluctuations is obtained. Can be reduced.

  FIG. 5a shows a received power distribution generated by extracting some distance ranges (X1 to X2) from a plurality of distance spectra including the distance spectrum shown in FIG. 4b and the ratio of the received power of the extracted distance spectrum. FIG. X1 and X2 are arbitrary values on the distance R axis and can be set in advance.

  The number of measurement cycles described above is predetermined, but taking into account the measurement conditions of each distance spectrum, if the vehicle has traveled a predetermined distance, even if the measurement cycle is less than the predetermined number, One received power distribution can be generated using the distance spectrum generated up to that point. In this case, the obstacle detection device 1 includes a vehicle speed sensor that can detect the speed of the host vehicle, and can measure the travel distance of the host vehicle.

  In the generation of the received power distribution described above, the distance range X1 to X2 can be the entire distance R axis. It is also possible to generate a received power distribution from a ratio of received power in a partial distance range of one distance spectrum (not a plurality).

  Regarding the generation of the received power distribution, the average received power in each distance spectrum can be extracted, and the received power distribution can be generated from the extracted average received power distribution. Alternatively, for some distance spectra, the received power at each distance is extracted for each distance spectrum, and for some other distance spectra, the average received power of each distance spectrum is extracted, and these data are A received power distribution can also be generated based on this. Furthermore, it is possible to generate a received power distribution by fitting a part of the distance range of the distance spectrum to the generated distance spectrum by a peak shape function or kernel density estimation and using the value of the fitted function.

  However, in the case of an on-vehicle radar, it is known that the received power characteristics fluctuate greatly even if the traveling object conditions are different even if the same object and the same distance are used. Therefore, for example, a signal from an object may be buried in a clutter (noise) signal and cannot be detected, or a clutter signal may be erroneously detected as an object.

  FIG. 5b shows the received power distribution representing the radar detection performance considering the variation of the received power intensity. Signal A (clutter signal) is the probability density of the received power when there is no object (target), and signal B (target signal + clutter signal) is the probability density when the object exists. Here, the detection probability S1 is a probability that the target object can be correctly detected when the target object exists, and the false alarm probability S2 is a probability that the target object is erroneously detected even if the target object does not exist. The detection probability and the false alarm probability vary depending on where the threshold value, which is a design value, is set. SCR (Signal to Clutter Ratio) is the ratio of the target signal intensity to the clutter signal intensity.

  The present inventor found that such noise varies depending on the driving environment (crossroads, T-shaped roads, presence / absence of buildings, presence / absence of parked vehicles, etc.). On the other hand, if the driving environment is the same or similar, the noise is also the same. Or it was found to be similar. In the present embodiment, the filter unit 44 determines a filter for removing the received power distribution caused by the traveling environment according to the traveling environment acquired by the traveling environment acquisition unit 41 in order to remove the above-described noise, and determines The received power distribution is filtered by the filtered filter.

Here, the filter determination method of the filter means 44 will be described. The filter according to the present embodiment is applied only to the received power distribution generated from the ratio of the received power of the distance spectrum as described above, that is, to the clutter (noise) component, in each traveling environment when no obstacle (object) exists. It is a filter that removes the resulting received power distribution (eg, signal A in FIG. 5b). The filter is preferably generated based on the received power distribution generated when no obstacle exists for each traveling environment. Filter, for example a crossroads filter _{F 1A, F 1B ..., F} 1N, T -junction filter F _2A, F _2B ..., can be F _2N, etc. The filter database stores, for each traveling environment, a filter that removes the received power distribution due to only the clutter component in each traveling environment as described above.

  For example, the filter unit 44 refers to the filter database using the travel environment acquired by the travel environment acquisition unit 41 as a query, thereby acquiring a filter corresponding to the travel environment related to the host vehicle, and performs a filtering process using the filter.

  Here, the filter that removes the received power distribution caused only by the clutter component is generated by specifying a known distribution that approximates the received power distribution. The known distribution can be various distributions such as a Rayleigh distribution, a Weibull distribution, a logistic distribution, a normal distribution, a Laplace distribution, a Cauchy distribution, and the distribution is specified by further changing the coefficient depending on the distribution. Can do.

  The traveling environment can be classified into a plurality of categories depending on, for example, crossroads, T-shaped roads, presence of buildings, presence of parked vehicles, or combinations thereof. Furthermore, the traveling environment can be classified in more detail, for example, by further categorizing a specific intersection or a school periphery. The filter database is stored for each traveling environment classified in this way. The traveling environment acquisition unit 41 acquires information about which category the traveling environment relating to the host vehicle belongs.

  As one example, the traveling environment acquisition unit 41 acquires a traveling environment related to the host vehicle from road parameters, building parameters, and parked vehicle parameters. The road parameter is a parameter indicating the shape of the road (straight line or curve, number of lanes, road width), the number of intersecting roads, the intersection angle, and the like. The building parameter is a parameter indicating the presence or absence of a building and the position and height of the building when there is a building. The parked vehicle parameter is a parameter indicating the presence or absence of a parked vehicle and the position or shape of the parked vehicle when there is a parked vehicle.

  In this case, a filter is generated for each combination of the parameters, and the filter database stores a filter for each combination of the parameters. The filter unit 44 acquires (determines) a filter corresponding to the various parameters by referring to the filter database for the various parameters acquired by the traveling environment acquisition unit 41. In this way, the filter unit 44 determines a filter according to the traveling environment related to the host vehicle, and executes a filtering process using the filter.

  Note that instead of the storage means 42 storing the filter database, the server can also store the filter database. In this case, the filter means 44 communicates with the server via the communication unit so that the filter corresponding to the traveling environment is stored. It is possible to obtain and perform filtering using the filter.

  The obstacle detection unit 45 detects an obstacle based on the received power distribution subjected to the filtering process. For example, it is determined whether or not the average power in the received power distribution is equal to or greater than a predetermined threshold, and it is determined that an obstacle exists in the corresponding distance range when the average power is equal to or greater than the threshold.

  As another embodiment, the analysis unit 43 extracts a signal for each predetermined distance range on the distance spectrum, and calculates the ratio of the received power of the distance spectrum within the extracted predetermined distance range for each predetermined distance range. The received power distribution is generated. Thus, by performing processing for each distance range from the host vehicle, the accuracy of obstacle detection can be further increased.

  As shown in FIG. 4c, the analysis unit 43 divides the distance spectrum into a plurality of predetermined distance ranges (X0 to X1, X1 to X2, X2 to X3, X3 to X4), and the plurality of predetermined distance ranges. The signal in each is extracted. In this case, a received power distribution is generated for each predetermined distance range. For example, the predetermined distance range can be set to 5 m to 10 m, 10 m to 15 m,..., 95 to 100 m, or can be set to 10 to 25 m, 25 to 40 m,. The distance range width of the plurality of predetermined distance ranges extracted by the analysis unit 43 is not limited to the above example, and can be appropriately set according to the performance of the radar. The generation of the received power distribution from the distance spectrum is as described above.

  The filter according to the embodiment includes a received power distribution generated from a ratio of received power of a distance spectrum for each of a plurality of predetermined distance ranges in each traveling environment when no obstacle (object) exists, that is, clutter. This is a filter that removes the received power distribution caused by only the (noise) component. The filter database stores a filter for removing the received power distribution caused by only the clutter component in each traveling environment as described above for each traveling environment and for each predetermined distance range. Note that the filter database may be configured to store the filter for each traveling environment and to deform the filter according to a predetermined distance range.

  In yet another embodiment, an obstacle detection device 1 is provided that takes into account the fluctuation characteristics that change from moment to moment in an actual in-vehicle running state.

  For example, if the acquired received power distribution data is data at a moment when the noise (clutter) signal is large and the object (target) signal is small, there is a possibility that the object cannot be detected correctly. FIG. 6 a shows a distance spectrum when the host vehicle is located at point A. At this time, it can be confirmed that there is a peak in the distance range X1 to X2. On the other hand, FIG. 6 b shows the received power distribution when the host vehicle is located at a point B moved by a minute distance from the point A. At this time, it is understood that it is difficult to confirm the peak in the distance range X1 to X2.

  The analysis means 43 according to this embodiment generates a distance spectrum as shown in FIGS. 6a and 6b by using an FMCW millimeter wave radar every time the host vehicle moves a very small distance. Then, the analysis unit 43 extracts a part of the distance range (for example, distances X1 to X2) from the generated plurality of distance spectra, and generates one received power distribution from the ratio of the received power of the extracted distance spectrum. . The generation of the received power distribution is as described above.

  The minute distance here is a distance that is minute relative to the distance from the host vehicle to the object and does not affect the received power in the radar equation. According to the embodiment, since the received power distribution is generated by changing the environmental factors such as multipath without changing the theoretical received power value, rather than using a plurality of distance spectra generated at the same position. It is possible to generate a received power distribution (and a filter that removes a received power distribution caused only by a clutter component) in consideration of a fluctuation characteristic that changes every moment in an actual in-vehicle running state. The minute distance can be a predetermined distance of 50 cm or less when the distance between the host vehicle and the object is 10 m, for example, and can be a predetermined distance of 1 m or less when the distance between the host vehicle and the object is 50 m. be able to.

  Next, a processing flow of the obstacle detection apparatus 1 of the present embodiment will be described based on FIG. This processing flow is repeated every predetermined number of measurement cycles.

  First, the analysis unit 43 generates a beat signal using an FMCW millimeter wave radar (S701). Subsequently, the analysis unit 43 generates a distance spectrum using a frequency spectrum generated by performing a complex FFT operation on the beat signal (S702).

  The analysis unit 43 determines whether or not a predetermined quantity of distance spectrum has been generated. If the predetermined quantity has been generated, the analysis unit 43 proceeds to the next step (S703). Subsequently, the analysis unit 43 generates a received power distribution using a predetermined number of distance spectra (S704).

  The filter unit 44 acquires the traveling environment acquired by the traveling environment acquisition unit 41 in a control process different from the present processing flow, and inquires of the traveling environment by referring to the filter database using the traveling environment. A corresponding filter is determined (acquired) (S705).

  In this process flow, the filter is determined using the travel environment acquired in another control process. However, the travel environment acquisition is a part of the control process of this flow, for example, filter determination (S705). The driving environment may be acquired by incorporating it immediately before or before the beat signal generation (S701).

  The filter unit 44 filters the received power distribution using the determined filter (S706). The obstacle detection unit 45 determines whether or not there is an obstacle by determining whether or not the average value of the filtered received power distribution is greater than or equal to a predetermined threshold (S707). In this way, this processing flow ends.

  Next, the effect of the obstacle detection device 1 according to the embodiment of the present invention will be described. In this embodiment, in the obstacle detection executed based on the received power distribution generated by the analysis unit 43 using the FMCW millimeter wave radar, the travel environment acquisition unit 41 acquires the travel environment related to the host vehicle, and the filter unit. The filter according to the driving environment acquired by 44 is determined, and the received power distribution is filtered by the determined filter. Thus, by performing filtering using a filter according to the traveling environment, clutter caused by the traveling environment can be reduced.

  Further, when the analysis unit 43 generates the received power distribution based on the generated plurality of distance spectra, the received power distribution is generated by averaging the temporal variation, and thus the temporal variation is reduced. A received power distribution can be generated.

  Further, when the analysis unit 43 generates a reception power distribution based on a plurality of distance spectra generated each time the host vehicle moves a minute distance, an environment such as multipath without changing the theoretical reception power value. Since the reception power distribution with different factors is generated, it is necessary to generate the reception power distribution that takes into account the fluctuation characteristics that change every moment in the actual in-vehicle driving state rather than using multiple distance spectra generated at the same position. Can do.

  Moreover, in this embodiment, the analysis means 43 extracts a signal for every predetermined distance range on the distance spectrum, and determines each predetermined distance range from the ratio of the received power of the distance spectrum within the extracted predetermined distance range. A reception power distribution for each is generated. Similarly, the filter subjected to the filtering process by the filter unit 44 is a filter generated for each of a plurality of predetermined distance ranges. Thus, by performing processing for each distance range from the host vehicle, the accuracy of obstacle detection can be further increased.

  Moreover, in this embodiment, the analysis means 43 produces | generates the received power distribution when an obstruction does not exist for every driving | running | working environment, and the filter which removes the said received power distribution is produced | generated based on the produced | generated received power distribution.

  Each Example described above is an illustration for explaining the present invention, and the present invention is not limited to these Examples. The present invention can be implemented in various forms without departing from the gist thereof.

DESCRIPTION OF SYMBOLS 1 Obstacle detection apparatus 10 Control part 11 Processing part 12 Storage part 13 Interface part 20 Radar 30 Camera 41 Running environment acquisition means 42 Storage means 43 Analysis means 44 Filter means 45 Obstacle detection means

Claims (5)

Driving environment acquisition means for acquiring a driving environment related to the host vehicle;
A received power distribution that generates a beat signal using an FMCW millimeter wave radar, generates a distance spectrum using a frequency spectrum derived from the beat signal, and indicates a probability density for each received power based on the distance spectrum Analyzing means for generating
Filter means for determining a filter for removing the received power distribution caused by the traveling environment according to the acquired traveling environment, and filtering the generated received power distribution by the determined filter;
An obstacle detection device comprising obstacle detection means for detecting an obstacle based on the filtered received power distribution.   The obstacle detection apparatus according to claim 1, wherein the analysis unit generates a received power distribution based on the plurality of generated distance spectra.   The obstacle detection device according to claim 1, wherein the analysis unit generates a received power distribution based on a plurality of the distance spectra generated each time the host vehicle moves a minute distance. The analysis means generates each received power distribution based on data within a plurality of predetermined distance ranges of the distance spectrum,
The filter means determines a filter for removing a received power distribution caused by a driving environment corresponding to each predetermined distance range according to the acquired driving environment, and the received power generated by the determined filter. The obstacle detection device according to any one of claims 1 to 3, wherein the distribution is filtered.   The obstacle detection device according to any one of claims 1 to 4, wherein the filter is generated based on a received power distribution generated by the analysis unit when there is no obstacle for each traveling environment.

Images