Method and system for detecting vehicle bottom obstacle by using ultrasonic sensor

Abstract

The invention relates to a method and system for detecting vehicle bottom obstacles using an ultrasonic sensor. The method comprises the following steps: acquiring an envelope curve of an ultrasonic signal through an ultrasonic sensor arranged at the bottom of a vehicle; performing feature extraction on the envelope curve to obtain sample points; calculating the distance between the sample point and the center point of the cluster; and determining whether there is an obstacle at the underbody based on a comparison of the distance to a predetermined threshold distance.

Description

Technical Field

[0001] The present invention relates to a method and a system for detecting obstacles under a vehicle using an ultrasonic sensor. Background Art

[0002] With the rapid development of the automobile industry and the continuous improvement of people's living standards, more and more cars have entered thousands of households. However, while cars provide people with travel convenience, they also bring many safety hazards. For example, the existence of blind spots under the car makes it difficult to avoid accidents caused by small animals or even children, because they may have entered the bottom of the car before the vehicle started, or they may enter the blind spots under the car while the vehicle is moving without being noticed or confirmed.

[0003] The present disclosure has been improved with respect to but not limited to the above-mentioned factors. Summary of the invention

[0004] To this end, the present disclosure proposes a method and system for monitoring the blind spot under the vehicle using an ultrasonic sensor. The method and system of the present disclosure use an ultrasonic sensor to collect the envelope curve of the ultrasonic signal, and use a clustering algorithm to distinguish whether there is an obstacle under the vehicle. If the measured envelope curve cannot be classified into any cluster of the clusters calibrated in advance under obstacle-free conditions, or the measured envelope curve is classified into any cluster of the clusters calibrated in advance under obstacle conditions, it can be determined that there is an obstacle under the vehicle, thereby achieving the purpose of detecting the blind spot under the vehicle.

[0005] According to a first aspect of the present disclosure, a method for detecting obstacles under a vehicle using an ultrasonic sensor is provided, comprising: collecting an envelope curve of an ultrasonic signal via an ultrasonic sensor disposed under the vehicle; performing feature extraction on the envelope curve to obtain sample points; calculating the distance between the sample points and a center point of a cluster; and determining whether there is an obstacle under the vehicle based on a comparison of the distance with a predetermined threshold distance.

[0006] According to an embodiment, the ultrasonic sensor is positioned to emit ultrasonic waves to cover the vehicle bottom area to the greatest extent.

[0007] According to another embodiment, the ultrasonic sensor includes at least two ultrasonic sensors, and two ultrasonic sensors of the at least two ultrasonic sensors are installed symmetrically with respect to a plane center in a vehicle bottom plane or parallel to the vehicle bottom plane.

[0008] According to yet another embodiment, one of the ultrasonic sensors is disposed in the middle of a front bumper of a vehicle and transmits ultrasonic waves toward a rear of the vehicle, and another of the ultrasonic sensors is disposed in the middle of a rear bumper of the vehicle and transmits ultrasonic waves toward a front of the vehicle.

[0009] According to another embodiment, collecting the envelope curve of the ultrasonic signal also includes: sequentially causing one of the at least two ultrasonic sensors to emit an ultrasonic wave so that the reflected ultrasonic signal is received by the other ultrasonic sensors among the at least two ultrasonic sensors, wherein receiving the reflected ultrasonic signal includes filtering out the non-reflected ultrasonic wave according to the distance between the ultrasonic sensors.

[0010] According to yet another embodiment, the ultrasonic sensor comprises a single ultrasonic sensor, and collecting the envelope curve of the ultrasonic signal further comprises switching the single ultrasonic sensor between a transmission mode and a reception mode so as to collect the envelope curve of the ultrasonic signal.

[0011] According to another embodiment, the clustering is obtained by clustering the envelope curves of the ultrasonic signals collected in a preset scenario with no obstacles under the vehicle in advance through a clustering algorithm, wherein there is one cluster for each preset scenario, wherein the preset scenario varies depending on at least one of the road condition, ambient wind speed and / or wind direction, and ambient noise level, wherein the method includes determining that there is an obstacle under the vehicle when the distance is greater than the predetermined threshold distance.

[0012] According to another embodiment, the clustering is obtained by clustering the envelope curves of the ultrasonic signals collected in the preset scenario where there is an obstacle under the vehicle in advance through a clustering algorithm, wherein there is one cluster for each preset scenario, wherein the preset scenario varies depending on at least one of the road condition, ambient wind speed and / or wind direction, and ambient noise level, wherein the method includes determining that there is an obstacle under the vehicle when the distance is less than the predetermined threshold distance.

[0013] According to yet another embodiment, the predetermined threshold distance varies from cluster to cluster, wherein the predetermined threshold distance is proportional to the range of the cluster.

[0014] According to yet another embodiment, the method is executed when the vehicle is started; or, the method is executed when an autonomous parking instruction is received.

[0015] According to yet another embodiment, the method further comprises ending the method after the speed of the vehicle exceeds a predetermined speed and / or the vehicle travels a predetermined distance; or ending the method after the vehicle is turned off.

[0016] According to yet another embodiment, the method further comprises updating the clustering using an envelope curve of the acquired ultrasonic signal.

[0017] According to yet another embodiment, the method further comprises issuing an alarm after determining that an obstacle exists under the vehicle.

[0018] According to a second aspect of the present disclosure, a system for detecting obstacles under a vehicle using an ultrasonic sensor is provided, comprising: an ultrasonic sensor disposed under the vehicle, wherein the ultrasonic sensor is configured to collect an envelope curve of an ultrasonic signal; and a computing device, wherein the computing device is configured to receive the envelope curve of the ultrasonic signal from the ultrasonic sensor, wherein the computing device is further configured to execute the method according to the first aspect of the present disclosure.

[0019] According to an embodiment, the computing device comprises an on-board computer and / or a cloud, and wherein in the case where the computing device comprises a cloud, the system further comprises a communication device for communicating with the cloud.

[0020] According to a third aspect of the present disclosure, a motor vehicle is provided, comprising the system according to the second aspect of the present disclosure.

[0021] Thus, compared with conventional sensors such as cameras and laser radars, the method and system disclosed in the present invention are small in size, low in price, not affected by lighting conditions, easy to deploy at the bottom of the vehicle, and highly feasible. And unlike conventional ultrasonic sensor obstacle detection methods, the method and system disclosed in the present invention does not simply use a threshold to determine the presence of an obstacle, but rather distinguishes whether there is an obstacle at the bottom of the vehicle by clustering the envelope curve, thereby improving the effectiveness, sensitivity, and robustness of obstacle detection.

[0022] Aspects generally include methods, apparatus, systems, computer program products, and processing systems substantially as described herein with reference to and as illustrated by the accompanying figures.

[0023] The foregoing has broadly outlined the features and technical advantages of examples according to the present disclosure so that the following detailed description can be better understood. Additional features and advantages will be described hereinafter. The disclosed concepts and specific examples can be easily used as the basis for modifying or designing other structures for implementing the same purpose as the present disclosure. Such equivalent constructions do not depart from the scope of the appended claims. The characteristics of the concepts disclosed herein, both in terms of their organization and method of operation, and the associated advantages will be better understood by considering the following description in conjunction with the accompanying drawings. Each of the accompanying drawings is provided for illustration and description purposes and does not define limitations on the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to understand in detail the manner in which the above-stated features of the present disclosure are used, the above briefly summarized content may be described in more detail with reference to various aspects, some of which are illustrated in the accompanying drawings. However, it should be noted that the accompanying drawings illustrate only certain typical aspects of the present disclosure and should not be considered to limit its scope, as the description may allow for other equally effective aspects. The same reference numerals in different drawings may identify the same or similar elements.

[0025] Figure 1 A flowchart showing an example method for detecting obstacles under a vehicle using an ultrasonic sensor according to an embodiment of the present disclosure;

[0026] Figure 2 A schematic diagram showing an envelope curve of an ultrasonic signal according to an embodiment of the present disclosure.

[0027] Figure 3 A schematic diagram showing the result of k clustering according to an example embodiment of the present disclosure.

[0028] Figure 4 A schematic block diagram showing a system for detecting obstacles under a vehicle using an ultrasonic sensor according to an example embodiment of the present disclosure; and

[0029] Figure 5 is a schematic diagram of an example motor vehicle according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0030] In recent years, more and more perception technologies and driver assistance systems have been applied to the automotive field. For example, 360-degree panoramic images can provide drivers with a view of the vehicle's surroundings, effectively eliminating blind spots. Autonomous parking technology uses multiple sets of sensors (including cameras, millimeter-wave radars, ultrasonic sensors, etc.) on the front, back and sides of the vehicle to sense the vehicle's surroundings and automatically complete parking.

[0031] At present, almost no technology for monitoring blind spots under the vehicle has been put into practical use. For the camera, the underside of the vehicle body is not well illuminated, which cannot provide a clear picture to the driver and is not conducive to the application of visual detection methods. Adding under-vehicle lighting will further increase the system cost. In addition, the bottom of the car is easily contaminated by mud and sand during driving, causing the camera to go blind. LiDAR and millimeter-wave radar are more expensive and larger in size, making them difficult to deploy under the vehicle body.

[0032] The inventors realized that ultrasonic sensors are small in size, low in price, use ultrasonic signals to sense the surrounding environment, have a large FOV, can effectively cover the bottom area of ​​the vehicle, and are suitable for blind spot monitoring tasks under the vehicle.

[0033] The inventors also realized that when the ultrasonic sensor is used as a reversing radar, the obstacle distance is determined by detecting the echo of the obstacle, that is, when the echo amplitude is higher than the set threshold, it is considered that there is an obstacle at the distance corresponding to the echo. If the ultrasonic sensor is simply arranged at the bottom of the vehicle and a similar method is used for monitoring, the effect is unsatisfactory. For example, when the ultrasonic sensor is arranged at the bottom of the vehicle body, it can be foreseen that there will almost always be objects within the FOV range of the ultrasonic sensor, and correspondingly there will be higher echoes, such as echoes from the ground and wheels, and even echoes reflected back and forth between the ground and the bottom of the vehicle. In addition, in different scenarios (including different road conditions, different ambient wind speeds and / or wind directions, different ambient noise levels, etc.), the ultrasonic echo conditions are different, and a simple threshold judgment cannot effectively distinguish whether there is really an obstacle. Even if calibration is performed, the sensitivity and robustness of the detection will be severely limited.

[0034] With the upgrade and iteration of ultrasonic sensors, the output of ultrasonic sensors is no longer limited to IO signals. For example, the ELMOS 524.09 chip can output the intensity value and corresponding distance of the obstacle echo. The new generation of ultrasonic sensors (such as the ELMOS 524.17 chip) can output the envelope curve of the ultrasonic signal, which contains richer raw data information. For example, the features corresponding to each scene can be extracted from the envelope curve. As a result, the inventors realized that this envelope curve is the basis for classification using machine learning algorithms (such as clustering algorithms).

[0035] To this end, the present disclosure proposes a method and system for monitoring the blind spot under the vehicle using an ultrasonic sensor. The method and system of the present disclosure use at least one ultrasonic sensor to transmit an ultrasonic signal to the bottom of the vehicle, and the received reflected ultrasonic signal is output to the processing device in the form of an envelope curve, and a clustering algorithm is used to distinguish whether there is an obstacle under the vehicle. If the measured envelope curve cannot be classified into any cluster of the clusters calibrated in advance under obstacle-free conditions, it can be determined that there is an obstacle under the vehicle, thereby achieving the purpose of blind spot detection under the vehicle. Therefore, the method and system of the present disclosure clusters the envelope curves of ultrasonic signals in obstacle-free scenes, and determines whether there is an obstacle under the vehicle by judging the subordinate relationship of the new measured envelope curve to the existing clustering cluster. This conforms to the characteristics of the clustering algorithm and has sufficient theoretical support.

[0036] Thus, compared with conventional sensors such as cameras and laser radars, the method and system disclosed in the present invention are small in size, low in price, not affected by lighting conditions, easy to deploy at the bottom of the vehicle, and highly feasible. And unlike conventional ultrasonic sensor obstacle detection methods, the method and system disclosed in the present invention does not simply use a threshold to determine the presence of an obstacle, but rather distinguishes whether there is an obstacle at the bottom of the vehicle by clustering the envelope curve, thereby improving the effectiveness, sensitivity, and robustness of obstacle detection.

[0037] The detailed description set forth below in conjunction with the accompanying drawings is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein may be practiced. This detailed description includes specific details to provide a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts may be practiced without these specific details.

[0038] refer to Figure 1 , which shows a flowchart of an example method 100 for detecting obstacles under a vehicle using an ultrasonic sensor according to an embodiment of the present disclosure.

[0039] As shown, the method 100 may include, at block 110 , collecting an envelope curve of an ultrasonic signal via an ultrasonic sensor disposed under the vehicle.

[0040] In one embodiment of the present disclosure, the ultrasonic sensor may be positioned to emit ultrasonic waves to cover the area under the vehicle to the greatest extent. For example, the ultrasonic sensor may be installed at an angle relative to the vehicle based on its field of view (FOV), installation location, vehicle model (including vehicle length, vehicle width, height of the vehicle bottom from the ground, etc.) and the like to cover the area under the vehicle to the greatest extent.

[0041] In one embodiment of the present disclosure, the ultrasonic sensor includes a single ultrasonic sensor, and the single ultrasonic sensor can be installed in the middle of the front bumper of the vehicle and transmit ultrasonic waves toward the rear of the vehicle, or installed in the middle of the rear bumper of the vehicle and transmit ultrasonic waves toward the front of the vehicle, so that the transmitted ultrasonic waves cover the entire bottom area of ​​the vehicle as much as possible. Further according to this embodiment, collecting the envelope curve of the ultrasonic signal may also include: switching the single ultrasonic sensor between a transmitting mode and a receiving mode to collect the envelope curve of the ultrasonic signal. For example, the single ultrasonic sensor may first transmit ultrasonic waves in a transmitting mode, and then switch to a receiving mode to receive the reflected ultrasonic signal.

[0042] In a preferred embodiment of the present disclosure, the ultrasonic sensor includes at least two ultrasonic sensors, and two of the at least two ultrasonic sensors are installed symmetrically relative to the center of the plane in the bottom plane of the vehicle or parallel to the bottom plane of the vehicle. For example, one of the ultrasonic sensors is disposed in the middle of the front bumper of the vehicle and transmits ultrasonic waves toward the rear of the vehicle, and another of the ultrasonic sensors is disposed in the middle of the rear bumper of the vehicle and transmits ultrasonic waves toward the front of the vehicle, so that the two ultrasonic sensors can cover the entire bottom area of ​​the vehicle. Further according to this embodiment, collecting the envelope curve of the ultrasonic signal may also include: sequentially causing one of the at least two ultrasonic sensors to transmit ultrasonic waves so that the reflected ultrasonic signals are received by the other of the at least two ultrasonic sensors.

[0043] For example, when two ultrasonic sensors are included, when one of the two ultrasonic sensors is configured to transmit ultrasonic waves and the other of the two ultrasonic sensors is configured to receive ultrasonic waves, the ultrasonic waves transmitted by the ultrasonic sensor configured to transmit ultrasonic waves may not be reflected by the ground and directly received by the ultrasonic sensor configured to receive ultrasonic waves. Such unreflected ultrasonic waves are invalid in the blind spot monitoring under the vehicle (or in other words, the obstacle detection under the vehicle), and should therefore be excluded. The inventors recognize that the propagation distance of the unreflected ultrasonic waves received by the ultrasonic sensor configured to receive ultrasonic waves is the straight-line distance between the ultrasonic sensor configured to transmit ultrasonic waves and the ultrasonic sensor configured to receive ultrasonic waves, and the propagation distance of the reflected ultrasonic waves is greater than this straight-line distance, so the propagation time of the reflected ultrasonic waves is greater than the propagation time of the unreflected ultrasonic waves. Thus, receiving the reflected ultrasonic signal may include filtering out the unreflected ultrasonic waves according to the distance between the ultrasonic sensors. Of course, in the filtering process, the internal delay of the ultrasonic sensor, the processing time of the ultrasonic signal, etc. may also be taken into account, which will not be repeated here.

[0044] For another example, when there are three ultrasonic sensors (e.g., the first, second, and third ultrasonic sensors), sequentially causing one of at least two ultrasonic sensors to transmit an ultrasonic wave so that the other ultrasonic sensors of the at least two ultrasonic sensors receive the reflected ultrasonic signal may include:

[0045] First, the first ultrasonic sensor transmits ultrasonic waves, and the second and third ultrasonic sensors receive reflected ultrasonic waves;

[0046] Secondly, the second ultrasonic sensor transmits ultrasonic waves, and the first and third ultrasonic sensors receive reflected ultrasonic waves; and

[0047] Finally, the third ultrasonic sensor transmits ultrasonic waves, and the first and second ultrasonic sensors receive reflected ultrasonic waves.

[0048] Of course, in this example, the unreflected ultrasonic waves received by the ultrasonic sensor as the receiver can also be filtered out as described above, which will not be described in detail here.

[0049] Continue to refer Figure 1 The method 100 may include, at block 120 , performing feature extraction on the envelope curve to obtain sample points, and at block 130 , calculating the distance between the sample points and the center point of the cluster.

[0050] In one embodiment of the present disclosure, clustering is obtained by clustering the envelope curves of ultrasonic signals collected in preset scenarios without obstacles under the vehicle in advance through a clustering algorithm, wherein each preset scenario has one cluster, and wherein the preset scenarios vary according to at least one of the road conditions, ambient wind speed and / or wind direction, and ambient noise level. In one example, first, in different scenarios where there are no obstacles under the vehicle (e.g., gravel road with no wind, gravel road with wind force level 1, asphalt road with wind force level 3, etc.), multiple sets of envelope curves of ultrasonic signals are collected. For example, see Figure 2 , which shows a schematic diagram of an envelope curve of an ultrasonic signal according to an exemplary embodiment of the present disclosure. Figure 2 As can be seen in the figure, the horizontal axis represents the distance from the ultrasonic sensor, and the vertical axis represents the echo intensity (dimensionless) of the reflected ultrasonic signal. Subsequently, the envelope curve is feature extracted to obtain sample points, i.e., feature vectors. It will be understood that the envelope curve exhibits different characteristics due to different conditions under the vehicle, such as the number of echoes, average echo intensity, echo intensity variance, etc. These features can be extracted from the envelope curve as sample points for clustering. Next, the set of sample points is subjected to k-means clustering. The k value may depend on the number of scenes, or it may be manually set in the actual clustering process depending on the data distribution. In this way, k cluster centers and the corresponding ranges of each cluster may be obtained. These clusters may then be used in step 130 of method 100 to calculate the distance between the sample point and the cluster center. It will be understood that the "distance" described in the present disclosure may be any suitable distance metric, such as Euclidean distance.

[0051] In another embodiment of the present disclosure, the envelope curve of the ultrasonic signal can be collected in the scene where there is an obstacle under the vehicle and clustered, so that these clusters can also be used for obstacle detection under the vehicle. Thus, the clustering can include the clustering of the envelope curve of the ultrasonic signal collected in the scene where there is no obstacle under the vehicle and the scene where there is an obstacle under the vehicle.

[0052] In one embodiment of the present disclosure, the clusters obtained in various scenarios can be summarized into cluster clusters and stored in the vehicle's configuration file to facilitate reading these parameters when the vehicle is running. Thus, step 130 of method 100 can be performed by the vehicle's on-board computer. In another embodiment of the present disclosure, the clusters obtained in various scenarios can be stored in the cloud. In this embodiment, method 100 may also include transmitting the collected envelope curves to the cloud so that steps 120 and 130 of method 100 are performed by the cloud.

[0053] See also Figure 3 , which shows a schematic diagram of the result of k clustering according to an exemplary embodiment of the present disclosure. Figure 3 It can be seen that in this example, the sample points are divided into 4 clusters, and the k value is 4; the range of each cluster is determined by the coordinates of the cluster center point and the cluster radius, such as Figure 3 In this example, the distance between the farthest sample point of the cluster and the cluster center can be used as the cluster radius, so that if the distance between the sample point and the cluster center is less than the cluster radius, it can be considered that the sample point belongs to the cluster.

[0054] Continue to refer Figure 1 , the method 100 may include, at block 140, determining whether there is an obstacle under the vehicle based on a comparison of the distance with a predetermined threshold distance. For example, when it is determined that the distance between the sample point and the center point of the cluster obtained in a preset scenario where there is no obstacle under the vehicle is greater than the predetermined threshold distance, it is determined that there is an obstacle under the vehicle. Figure 3 , if the sample points are determined to be Figure 3 The distance between each of the centers of the four clusters shown is greater than the corresponding cluster radius, so it can be considered that the sample point does not belong to any of the four clusters. Since these four clusters are generated in a scene without obstacles under the vehicle, it can be considered that the sample point indicates the presence of an obstacle under the vehicle.

[0055] For another example, when it is determined that the distance between the sample point and the center point of the cluster obtained in the preset scene where there is an obstacle under the vehicle is less than a predetermined threshold distance, it is determined that there is an obstacle under the vehicle. For example, if it is determined that the distance between the sample point and the center point of a cluster obtained in the preset scene where there is an obstacle under the vehicle is less than the radius of the cluster, it can be considered that the sample point belongs to the cluster. Since the cluster is generated in the scene where there is an obstacle under the vehicle, it can be considered that the sample point indicates that there is an obstacle under the vehicle.

[0056] It will be appreciated that the predetermined threshold distance may be any suitable value. However, in a preferred embodiment of the present disclosure, the predetermined threshold distance may be cluster-specific and may be proportional to the size of the cluster. For example, for clusters with a larger size, the corresponding predetermined threshold distance may be larger. Figure 3 , for the two clusters on the right, the predetermined threshold distance corresponding to the upper cluster may be greater than the predetermined threshold distance corresponding to the lower cluster.

[0057] It will be understood that the envelope curve of the ultrasonic signal is different for the transmitter ultrasonic sensor and the receiver ultrasonic sensor. In this way, the envelope curves of the ultrasonic signals collected in the case of different transmitter ultrasonic sensors or different receiver ultrasonic sensors are not used in the same clustering process, so that the clustering corresponds to a specific pair of transmitter and receiver ultrasonic sensors. Therefore, the method 100 may also optionally include identifying the transmitter and receiver ultrasonic sensor pairs of the collected ultrasonic signals, and selecting the corresponding cluster based on the ultrasonic sensor pair to perform steps 130 and 140, without performing these steps for all clusters. Of course, in the case of using only one ultrasonic sensor, this identification operation is not required.

[0058] In another preferred embodiment of the present disclosure, the envelope curve of the collected ultrasonic signal is not compared with all clusters, but only compared with the filtered clusters. For example, considering that the vehicle may frequently park in a limited number of parking locations (e.g., the driver's workplace, residence, etc.), the method 100 may also optionally include collecting the location information of the vehicle, and filtering the clusters corresponding to the location information based on it, so that only the distance between the sample point obtained by feature extraction of the envelope curve of the collected ultrasonic signal and the center of the filtered cluster can be calculated, and compared with the corresponding predetermined distance threshold.

[0059] In a preferred embodiment of the present disclosure, method 100 may be started when the vehicle is started. For example, method 100 may be started when the vehicle is started from an ignition-off state. Thus, method 100 may include first detecting the start of the vehicle (e.g., the start button is turned on, the engine is ignited, etc.) and then executing steps 110-140. In this embodiment, method 100 may end after the speed of the vehicle exceeds a predetermined speed and / or the vehicle travels a predetermined distance. For example, after the vehicle has traveled a distance equal to the length of the vehicle, it can be considered that the entire bottom of the vehicle has left the original parking position, so there is no need to detect obstacles under the vehicle, and thus method 100 can end.

[0060] In another preferred embodiment of the present disclosure, the method 100 may start executing when an autonomous parking instruction is received. For example, after the driver presses the autonomous parking button, the method 100 may start executing. Thus, the method 100 may include first detecting whether an autonomous parking instruction is received, and executing steps 110-140 after receiving the autonomous parking instruction. In this embodiment, the method 100 may end after the vehicle is turned off. For example, after the vehicle is turned off, it can be considered that the vehicle has been parked, and thus the method 100 may end.

[0061] In a preferred embodiment of the present disclosure, clustering can be updated, for example, using the envelope curve of the collected ultrasonic signal. For example, after the method 100 ends, the envelope curve of the collected ultrasonic signal can be used to update the pre-formed clustering. In one embodiment of the present disclosure, this update can be performed by an on-board computer, and the updated clustering can be stored in an on-board memory (e.g., in an on-board system configuration file) for subsequent use. In another embodiment of the present disclosure, this update can be performed by the cloud. In this embodiment, the envelope curve of the collected ultrasonic signal can be transmitted to the cloud to perform the update of the clustering by the cloud, and then the updated clustering can be transmitted back to the vehicle. It will be understood that this transmission can be incremental (i.e., only the increment (or difference) between the updated clustering and the original clustering is transmitted, rather than the entire updated clustering) to save network resources.

[0062] In another embodiment of the present disclosure, steps 120-140 of method 100 may be performed in the cloud, whereby method 100 may include transmitting the envelope curve of the collected ultrasonic signal to the cloud for the cloud to perform steps 120-140. Subsequently, method 100 may include transmitting the result of determining whether there is an obstacle under the vehicle to the vehicle. Further according to this embodiment, method 100 may notify the vehicle only when there is an obstacle under the vehicle to save network resources.

[0063] In another embodiment of the present disclosure, the method 100 may further include, after determining that there is an obstacle under the vehicle, sounding an alarm. In this embodiment, the alarm may take any suitable form, such as audio, light, vibration, etc. The driver may take any suitable measures after noticing the alarm, such as braking the vehicle, getting off the vehicle to check, etc. Further according to this embodiment, the driver may give feedback to the alarm, for example, in the case of determining a false alarm, the driver may cancel the alarm. Thus, the method 100 may receive this feedback and update the cluster accordingly.

[0064] Reference below Figure 4 , which shows a schematic block diagram of a system 400 for detecting obstacles under a vehicle using an ultrasonic sensor according to an example embodiment of the present disclosure.

[0065] As shown in the figure, the system 400 may include an ultrasonic sensor 401 disposed under the vehicle and a computing device 403. In an embodiment of the present disclosure, the ultrasonic sensor 401 may be configured to collect an envelope curve of an ultrasonic signal and transmit it to the computing device 403, and the computing device 403 may be configured to receive the envelope curve of the ultrasonic signal from the ultrasonic sensor 401 and execute the methods described in various embodiments of the present disclosure, such as the method 100.

[0066] In a preferred embodiment of the present disclosure, the computing device 403 may include an onboard computer and / or a cloud, so that the method of the present disclosure (such as method 100) can be executed on the onboard computer and the cloud in any suitable manner. For example, steps 120-140 of method 100 can be executed entirely by the cloud, entirely by the onboard computer, or partially on the cloud and partially on the onboard computer. In the case where the computing device 403 includes a cloud, the system 400 may also include a communication device 405 for communicating with the cloud, such as Figure 4 As shown in the dashed box 405 in FIG.

[0067] will understand, although Figure 4 Only one ultrasonic sensor 401 is shown, but the system 400 may include any suitable number of ultrasonic sensors, such as Figure 4 As shown by ellipsis 407 in FIG.

[0068] Figure 5 FIG. 5 is a schematic diagram of a motor vehicle 500 according to an exemplary embodiment of the present disclosure. In this embodiment, the motor vehicle 500 may include reference Figure 4 An example system 400 is shown and described.

[0069] As described above, the method and system of the present disclosure transmit and receive ultrasonic signals by installing at least one ultrasonic sensor at the bottom of the vehicle for monitoring the blind spot under the vehicle. The orientation of the ultrasonic sensor can be selected according to the actual situation to cover the bottom area of ​​the vehicle to the greatest extent, for example, the probe of the ultrasonic sensor is oriented parallel to the vehicle chassis and slightly tilted downward. Preferably, multiple ultrasonic sensors can be installed under the vehicle to achieve better blind spot monitoring effects. For example, an ultrasonic sensor can be installed near each of the four wheels, and each sensor can be self-transmitting and self-receiving or can be sent and received multiple times to increase the detection capability. In the case where there are no obstacles under the vehicle, a certain number of measurement curves (i.e., envelope curves of ultrasonic signals) are collected in advance for clustering. The clustering method can use algorithms such as k-Means (k-means), DBSCAN, etc. The center point of each cluster is represented by the mean (center of mass) of all sample points in each cluster. In order to improve robustness, test curve collection and clustering operations can be performed in different scenarios, such as windy or windless conditions, asphalt roads or gravel roads, and so on. Finally, the clustering results under different conditions can be summarized into clusters as the calibration results, which can be used to compare with the envelope curve collected in real time to detect obstacles under the vehicle. The clustering results under each scenario can be summarized into clusters and can be stored in the vehicle parking system configuration file, which is convenient for direct reading and use when the vehicle is running. The above calibration process can be completed before the vehicle leaves the factory as a common parameter of the blind spot monitoring system under the vehicle of the same model. In addition, the method and system disclosed in the present invention can also perform clustering online, for example, when the vehicle is about to enter the parking space, or within one or two minutes after the vehicle enters the parking space, the envelope curve of the ultrasonic signal is collected. At this time, since the vehicle can complete the parking action, it means that there is no obstacle under the vehicle. The envelope curve obtained in this way can be considered as a calibrated result, that is, a new clustering result. This new cluster can be added to the original cluster cluster and used for blind spot monitoring under the vehicle when the vehicle is started and prepared to leave the parking space next time.

[0070] The method and system disclosed in the present invention apply ultrasonic sensors to the task of detecting the blind spot under the vehicle, avoiding the problem that other sensors (such as cameras, laser radars, etc.) are difficult to deploy in this scenario, and the cost is low, and a large range of blind spots under the vehicle can be covered. In addition, the method and system disclosed in the present invention abandon the conventional method of using the ultrasonic sensor signal strength threshold to distinguish obstacles, and instead use a clustering method to comprehensively consider the ultrasonic signal envelope curves in different obstacle-free scenarios during the calibration stage, so as to effectively cover the dynamic scenes of target detection in the blind spot monitoring system under the vehicle, and achieve a more robust and more reliable detection effect. Therefore, the method and system disclosed in the present invention are highly flexible, and can not only be uniformly calibrated before the vehicle leaves the factory, but also can be calibrated online, incrementally calibrated, and so on during the use of the vehicle. The results of the online calibration (i.e., the new clustering results) can be flexibly added to the original clustering clusters, so as to achieve better detection results that are more adapted to the on-site situation.

[0071] Furthermore, those skilled in the art will appreciate that, although the present disclosure describes various embodiments in conjunction with clustering methods, any other suitable machine learning methods may be used, such as naive Bayes classifiers, logistic regression, decision trees, support vector machines, random forests, K-nearest neighbors, deep learning, neural networks, and the like.

[0072] Those skilled in the art will appreciate that, although various embodiments are described in conjunction with a motor vehicle in the present disclosure, the methods and systems of the present disclosure may also be applied to other suitable vehicles, such as airplanes, ships, etc., with corresponding adaptive modifications.

[0073] Those skilled in the art will appreciate that references to “the envelope curve of an ultrasonic signal” in this disclosure refer to the envelope curve of an ultrasonic signal reflected by the ground.

[0074] The above detailed description includes references to the accompanying drawings, which form part of the detailed description. The accompanying drawings illustrate specific embodiments that can be practiced by way of illustration. These embodiments are also referred to herein as "examples". Such examples may include elements other than those shown or described. However, examples including the elements shown or described are also contemplated. In addition, examples using any combination or arrangement of those elements shown or described are also contemplated, or with reference to the specific examples (or one or more aspects thereof) shown or described herein, or with reference to other examples (or one or more aspects thereof) shown or described herein.

[0075] In the appended claims, the terms "including" and "comprising" are open ended, that is, systems, devices, articles, or processes having elements other than those elements listed after such terms in a claim are still deemed to fall within the scope of that claim. Furthermore, in the appended claims, the terms "first," "second," and "third," etc. are used merely as labels and are not intended to indicate a numerical order to their objects.

[0076] In addition, the order of each operation explained in this specification is exemplary. In alternative embodiments, each operation can be performed in a different order than that shown in the drawings, and each operation can be combined into a single operation or split into more operations.

[0077] The above description is intended to be illustrative, not restrictive. For example, the examples described above (or one or more aspects thereof) may be used in conjunction with other embodiments. Other embodiments may be used, such as by a person of ordinary skill in the art, after reviewing the above description. The abstract allows the reader to quickly determine the nature of the present technology disclosure. The abstract is submitted, and it is understood that the abstract will not be used to interpret or limit the scope or meaning of the claims. In addition, in the above specific embodiments, various features may be grouped together to make the disclosure fluent. However, the claims may not state every feature disclosed herein, because the embodiments may characterize a subset of the features. In addition, the embodiments may include fewer features than those disclosed in a particular example. Therefore, the attached claims are thus incorporated into the specific embodiments, and a claim exists independently as a separate embodiment. The scope of the embodiments disclosed herein should be determined with reference to the attached claims and the full scope of equivalents to which such claims are entitled.

Claims

1. A method for detecting an obstacle under a vehicle using an ultrasonic sensor, comprising: The envelope curve of the ultrasonic signal is collected by an ultrasonic sensor arranged at the bottom of the vehicle; Performing feature extraction on the envelope curve to obtain sample points; Calculating the distance between the sample point and the center point of the cluster; as well as Whether there is an obstacle under the vehicle is determined based on a comparison between the distance and a predetermined threshold distance.

2. The method according to claim 1, characterized in that: The ultrasonic sensors are positioned to emit ultrasonic waves to cover the vehicle bottom area to the greatest extent possible.

3. The method according to claim 2, characterized in that The ultrasonic sensor includes at least two ultrasonic sensors, and two ultrasonic sensors of the at least two ultrasonic sensors are installed symmetrically with respect to a plane center in a vehicle bottom plane or parallel to the vehicle bottom plane.

4. The method according to claim 3, characterized in that: One of the ultrasonic sensors is disposed in the middle of a front bumper of a vehicle and transmits ultrasonic waves toward a rear of the vehicle, and the other of the ultrasonic sensors is disposed in the middle of a rear bumper of the vehicle and transmits ultrasonic waves toward a front of the vehicle.

5. The method according to claim 4, characterized in that The envelope curve of the collected ultrasonic signal also includes: sequentially causing one of the at least two ultrasonic sensors to transmit an ultrasonic wave so that the other ultrasonic sensor of the at least two ultrasonic sensors receives the reflected ultrasonic wave signal, Receiving the reflected ultrasonic signal includes filtering out the unreflected ultrasonic signal according to the distance between the ultrasonic sensors.

6. The method according to claim 2, characterized in that The ultrasonic sensor includes a single ultrasonic sensor, and collecting the envelope curve of the ultrasonic signal further includes switching the single ultrasonic sensor between a transmission mode and a reception mode so as to collect the envelope curve of the ultrasonic signal.

7. The method according to claim 1, characterized in that The clustering is obtained by clustering the envelope curves of ultrasonic signals collected in preset scenarios without obstacles under the vehicle in advance through a clustering algorithm, wherein each preset scenario has one cluster, wherein the preset scenarios vary depending on at least one of road conditions, ambient wind speed and / or wind direction, and ambient noise level, wherein the method includes determining that there is an obstacle under the vehicle when the distance is greater than the predetermined threshold distance.

8. The method according to claim 1, characterized in that: The clustering is obtained by clustering the envelope curves of ultrasonic signals collected in preset scenarios where there are obstacles under the vehicle in advance through a clustering algorithm, wherein each preset scenario has one cluster, wherein the preset scenarios vary depending on at least one of the road conditions, ambient wind speed and / or wind direction, and ambient noise level, wherein the method includes determining that there is an obstacle under the vehicle when the distance is less than the predetermined threshold distance.

9. The method according to claim 7 or 8, characterized in that: The predetermined threshold distance varies from cluster to cluster, wherein the predetermined threshold distance is proportional to the range of the cluster.

10. The method according to claim 1, characterized in that The method is executed when the vehicle is started; or, The method is executed upon receiving an autonomous parking instruction.

11. The method according to claim 10, characterized in that The method also includes ending after the speed of the vehicle exceeds a predetermined speed and / or the vehicle travels a predetermined distance; or, After the vehicle is turned off, the method ends.

12. The method according to any one of claims 1 to 8, characterized in that The method also includes updating the clustering using an envelope curve of the collected ultrasonic signal.

13. The method according to claim 1, characterized in that It also includes issuing an alarm after determining that there is an obstacle under the vehicle.

14. A system for detecting obstacles under a vehicle using an ultrasonic sensor, comprising: An ultrasonic sensor disposed under the vehicle, wherein the ultrasonic sensor is configured to collect an envelope curve of an ultrasonic signal; as well as a computing device configured to receive an envelope curve of the ultrasonic signal from the ultrasonic sensor, The computing device is further configured to execute the method as claimed in any one of claims 1-13.

15. The system according to claim 14, characterized in that The computing device comprises an on-board computer and / or a cloud, and wherein in the case where the computing device comprises a cloud, the system further comprises a communication device for communicating with the cloud.

16. A motor vehicle comprising a system according to any one of claims 14-15.

Images