Target detection method in automatic driving, electronic equipment and readable storage medium

Abstract

The invention discloses a target detection method in automatic driving, electronic equipment and a readable storage medium, and belongs to the technical field of automatic driving. The planning method comprises the following steps: adjusting detection ranges of a first sensor and a second sensor according to a driving scene and / or a vehicle body length; obtaining an initial probability of the target object indicated by first data of a first sensor and second data; if the association between the second data and the target object fails, determining a first prediction probability of the target object according to the adjusted detection range of the second sensor, the first data and the second data; and updating the initial probability according to the first prediction probability so as to determine whether the target object exists according to the updated initial probability. Through the method, the influence of the fluctuation of the detection range of the sensor on target detection can be fully considered, and the reliability and accuracy of the detection result are improved.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and particularly to a method, an electronic device, and a readable storage medium for object detection in autonomous driving. Background Art

[0002] Object detection in autonomous driving refers to the identification and positioning of potential obstacle areas, pedestrians, other vehicles, and other objects in the vehicle's surrounding environment during the autonomous driving process through various hardware (such as sensors like cameras, lidar, millimeter-wave radars, etc.) and software (such as algorithms). Object detection is a key component of autonomous driving technology and provides an important basis for subsequent path planning, decision-making, and control.

[0003] In related technologies, to improve the reliability and accuracy of object detection, the predicted value of the probability of the existence of an object by each sensor is usually calculated independently, and then it is determined whether the object exists based on the combined probability of existence, which can effectively reduce the impact of misidentification by individual sensors on the reliability and accuracy of object detection.

[0004] However, since the predicted value of each sensor in the above method affects the final result, when there is a reasonable missed detection situation in the sensor, it will also greatly reduce the probability of the existence of the object and affect the accuracy of the detection result. Summary of the Invention

[0005] A method, an electronic device, and a readable storage medium for object detection in autonomous driving are provided to solve the technical problems existing in related technologies. Specifically, the following technical solutions are included.

[0006] In a first aspect, the present application provides a method for object detection in autonomous driving. The method is used to detect target objects in the surrounding environment of a vehicle, and at least a first sensor and a second sensor are mounted on the vehicle. The method includes: adjusting the detection ranges of the first sensor and the second sensor according to the driving scenario and / or vehicle type; obtaining an initial probability of the target object indicated by first data and second data, where the first data corresponds to the first sensor after adjusting the detection range, and the second data corresponds to the second sensor after adjusting the detection range, and the initial probability is used to indicate the predicted value of the presence probability of the target object by the first sensor; if the second data fails to be associated with the target object, determining a first prediction probability of the target object according to the detection range of the second sensor after adjustment, the first data, and the second data, where the first prediction probability is used to indicate the predicted value of the presence probability of the target object by the second sensor when the second data fails to be associated with the target object; updating the initial probability according to the first prediction probability to determine whether the target object exists according to the updated initial probability.

[0007] In some possible implementation manners, adjusting the detection ranges of the first sensor and the second sensor according to the driving scenario and / or vehicle type includes: if the driving scenario is an overtaking scenario, reducing the detection ranges of the first sensor and the second sensor.

[0008] In some possible implementation manners, adjusting the detection ranges of the first sensor and the second sensor according to the driving scenario and / or vehicle type includes: if the vehicle type is a bus or a truck, reducing the detection ranges of the first sensor and the second sensor.

[0009] In some possible implementation manners, determining the first prediction probability of the target object according to the detection range of the second sensor after adjustment, the first data, and the second data includes: determining the position of the target object according to the first data; if the position is within the detection range of the second sensor after adjustment, determining the miss rate of the second sensor according to the relative relationship between the position and the detection range of the second sensor after adjustment; determining the first prediction probability according to the position, the miss rate, and the second data.

[0010] In some possible implementation manners, if the position is not within the detection range of the second sensor after adjustment, reducing the influence coefficient of the first prediction probability on the initial probability.

[0011] In some possible implementation manners, determining the first prediction probability according to the position, the missed detection rate, and the second data includes: determining an occlusion rate of the second sensor when detecting the target object according to the position and the second data; determining the first prediction probability according to the occlusion rate, the missed detection rate, and the second data.

[0012] In some possible implementation manners, the method further includes: if the second data is successfully associated with the target object, determining a second prediction probability of the target object according to the speed of the vehicle, the driving scenario, and the second data, where the second prediction probability is used to indicate a predicted value of the presence probability of the target object when the second data is successfully associated with the target object by the second sensor.

[0013] In some possible implementation manners, determining whether the target object exists according to the updated initial probability includes: if the updated initial probability is less than a deletion threshold, determining that the target object corresponding to the initial probability less than the deletion threshold does not exist; if the updated initial probability is greater than the deletion threshold, continuously updating the initial probability of the target object corresponding to the initial probability greater than the deletion threshold; if the updated initial probability is greater than a release threshold, determining that the target object corresponding to the initial probability greater than the release threshold exists, where the deletion threshold is less than the release threshold, the deletion threshold is used to indicate a critical value for the non - existence of the target object, and the release threshold is used to indicate a critical value for the existence of the target object.

[0014] In a second aspect, the present application further provides a control device for target detection in autonomous driving, including: an adjustment module, configured to adjust the detection ranges of the first sensor and the second sensor according to a driving scenario and / or a vehicle type. An acquisition module, configured to acquire an initial probability of the target object indicated by first data and second data, where the first data corresponds to the first sensor after adjusting the detection range, the second data corresponds to the second sensor after adjusting the detection range, and the initial probability is used to indicate a predicted value of the presence probability of the target object by the first sensor. A determination module, configured to, if the second data is not successfully associated with the target object, determine a first prediction probability of the target object according to the detection range of the second sensor after adjustment, the first data, and the second data, where the first prediction probability is used to indicate a predicted value of the presence probability of the target object by the second sensor when the second data is not successfully associated with the target object. An update module, configured to update the initial probability according to the first prediction probability, so as to determine whether the target object exists according to the updated initial probability.

[0015] In some possible embodiments, the adjustment module is configured to, if the driving scenario is an overtaking scenario, reduce the detection ranges of the first sensor and the second sensor.

[0016] In some possible embodiments, the adjustment module is configured to, if the vehicle type is a bus or a truck, reduce the detection ranges of the first sensor and the second sensor.

[0017] In some possible embodiments, the determination module is configured to determine the position of the target object according to the first data; if the position is within the detection range of the adjusted second sensor, determine the miss rate of the second sensor according to the relative relationship between the position and the detection range of the adjusted second sensor; and determine the first prediction probability according to the position, the miss rate, and the second data.

[0018] In some possible embodiments, the determination module is configured to, if the position is not within the detection range of the adjusted second sensor, reduce the influence coefficient of the first prediction probability on the initial probability.

[0019] In some possible embodiments, the determination module is configured to determine the occlusion rate of the second sensor when detecting the target object according to the position and the second data; and determine the first prediction probability according to the occlusion rate, the miss rate, and the second data.

[0020] In some possible embodiments, the determination module is further configured to, if the second data is successfully associated with the target object, determine a second prediction probability of the target object according to the speed of the vehicle, the driving scenario, and the second data, where the second prediction probability is used to indicate a predicted value of the existence probability of the target object when the second data is successfully associated with the target object by the second sensor.

[0021] In some possible embodiments, the update module is configured to, if the updated initial probability is less than the deletion threshold, determine that the target object corresponding to the initial probability less than the deletion threshold does not exist; if the updated initial probability is greater than the deletion threshold, continuously update the initial probability of the target object corresponding to the initial probability greater than the deletion threshold; and if the updated initial probability is greater than the release threshold, determine that the target object corresponding to the initial probability greater than the release threshold exists, where the deletion threshold is less than the release threshold, the deletion threshold is used to indicate a critical value for the non - existence of the target object, and the release threshold is used to indicate a critical value for the existence of the target object.

[0022] In a third aspect, the present application provides an electronic device for object detection in autonomous driving, including: a memory storing at least one program instruction for object detection in autonomous driving; a processor, when the program instruction is executed by the processor, enabling the vehicle to implement the method in the first aspect or any possible implementation manner of the first aspect of the present application.

[0023] In a fourth aspect, the present application provides a computer program (product), the computer program (product) includes computer programs / instructions, and when the computer programs / instructions are executed by a processor, enabling the vehicle to implement the method in the first aspect or any possible implementation manner of the first aspect of the present application.

[0024] In a fifth aspect, the present application provides a computer-readable storage medium storing program instructions for object detection in autonomous driving, and when the program instructions are executed by one or more processors, enabling the vehicle to implement the method in the first aspect or any possible implementation manner of the first aspect of the present application.

[0025] The beneficial effects of the technical solutions provided by the present application at least include:

[0026] The technical solutions disclosed in the present application fully consider the missed detection situations caused by the fluctuations in the detection ranges of sensors, and can timely adjust the detection ranges of the first sensor and the second sensor according to the driving scenarios of the vehicle and / or the vehicle type, so as to improve the accuracy and reliability of the detection results. For example, when the second data fails to be associated with the target object (that is, when the second sensor does not detect the target object), the first predicted value of the probability of the existence of the target object by the second sensor is determined according to the adjusted range, effectively avoiding the reduction of the probability of the existence of the target object caused by the missed detection situation due to the fluctuations in the detection ranges of the sensors, and making the detection results of object detection more reliable and accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0028] Figure 1 is a schematic diagram of an implementation scenario provided by an embodiment of the present application;

[0029] Figure 2 is a flowchart of the method for object detection in autonomous driving provided by the present application;

[0030] Figure 3 is a schematic diagram of the first confidence curve provided by an embodiment of the present application;

[0031] Figure 4 is a schematic structural diagram of a control device for object detection in autonomous driving provided by an embodiment of the present application;

[0032] Figure 5 is a schematic structural diagram of an electronic device for object detection in autonomous driving provided by an embodiment of the present application. Detailed implementation manners

[0033] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0034] Here, the exemplary embodiments will be described in detail, and the examples are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0035] Figure 1 is a schematic diagram of an implementation scenario provided by an embodiment of the present application. Refer to Figure 1 , the implementation scenario provided by the embodiment of the present application includes a vehicle 11 and an autonomous driving system 12.

[0036] The vehicle 11 is equipped with multiple sensors such as cameras, lidars, and millimeter-wave radars, which can be used to detect the surrounding environment of the vehicle 11. The autonomous driving system 12 can be connected to the multiple sensors mounted on the vehicle 11 in a wired or wireless manner, and can determine the probability of the existence of a target object in the surrounding environment of the vehicle 11 based on the multiple sensors mounted on the vehicle 11, and judge whether the target object exists.

[0037] Optionally, the autonomous driving system 12 can be a server, a server cluster, a cloud server, etc., and the present application does not make any restrictions on this.

[0038] Those skilled in the art should understand that the above-mentioned vehicle 11 and autonomous driving system 12 are only examples, and other existing or future vehicles and computer electronic devices applicable to the present application should also be included in the protection scope of the present application and are hereby incorporated by reference.

[0039] Figure 2It is a flowchart of a method for object detection in autonomous driving provided by this application. This method can be executed, for example, by a server, a server cluster, or a cloud server communicatively connected to an autonomous vehicle. This application does not impose any restrictions on this. Refer to Figure 2 The method for object detection in autonomous driving provided by the embodiments of this application can be used to detect potential target objects in the surrounding environment of the vehicle. At least a first sensor and a second sensor are mounted on the vehicle. This method can include the following steps.

[0040] Step S210, adjust the detection ranges of the first sensor and the second sensor according to the driving scenario and / or vehicle type.

[0041] Schematically, the driving scenario of the vehicle can be used to indicate the sum of the environment and conditions related to the driving of the vehicle, such as the motion state, weather, road, etc. of the vehicle; the vehicle type can include, for example, the distinction of the vehicle made according to the size, structure, and outer contour shape characteristics of the vehicle.

[0042] Considering that in the actual application scenario, the detection range of the sensor may fluctuate due to the influence of the driving scenario and vehicle type, which may cause the second sensor to fail to detect potential target objects in the surrounding environment of the vehicle detected by the first sensor. For example, when the second data is associated with the target object unsuccessfully, there may be two situations: one is that the first sensor misidentifies potential target objects in the surrounding environment of the vehicle; the other is that the detection range of the second sensor changes and fails to detect potential target objects in the surrounding environment of the vehicle. In view of this, the embodiments of this application can adjust the detection ranges of the first sensor and the second sensor according to the driving scenario and / or vehicle type to avoid the influence of the fluctuation of the detection range of the sensor on the detection result of the target object in the surrounding environment of the vehicle and improve the accuracy and reliability of the detection result.

[0043] In some embodiments, when the vehicle overtakes, the position and speed of the vehicle will change continuously, which may affect the actual detection range of the second sensor. In view of this, adjusting the detection ranges of the first sensor and the second sensor according to the driving scenario and / or vehicle type can include, for example: if the driving scenario of the vehicle is an overtaking scenario, reduce the detection ranges of the first sensor and the second sensor. Among them. The adjustment amplitude of the detection range of the first sensor / second sensor can be adjusted according to the actual application situation, and this application does not impose any restrictions on this.

[0044] In some embodiments, the method for identifying whether the driving scenario of the vehicle is an overtaking scenario can include, for example: if the lateral position of the vehicle changes, the left turn signal is turned on, and the speed difference in the longitudinal movement is greater than the speed difference threshold, it is determined that the driving scenario of the vehicle is an overtaking scenario. Among them, the value of the speed difference threshold can be adjusted according to the actual application situation, and this application does not impose any restrictions on this.

[0045] In some embodiments, due to the special body length, height, and outer contour shape of large vehicles such as buses and trucks, the actual detection ranges of the first sensor and the second sensor may be affected. In view of this, the detection ranges of the first sensor and the second sensor are adjusted according to the driving scenario and / or vehicle type. For example, it may include: if the vehicle type is a bus or a truck, the detection ranges of the first sensor and the second sensor are adjusted to be smaller. Similarly, the adjustment amplitude of the detection range of the first sensor / second sensor can be adjusted according to the actual application situation, and this application does not limit this.

[0046] In some embodiments, the method for determining the vehicle type may include, for example: determining the vehicle type according to the vehicle identification number. Among them, the vehicle identification number is used to indicate the unique identity of the vehicle and may include information such as the vehicle manufacturer, model, and production year.

[0047] Step S220, obtain the initial probability of the target object indicated by the first data and the second data. The first data corresponds to the first sensor after the detection range is adjusted, and the second data corresponds to the second sensor after the detection range is adjusted. The initial probability is used to indicate the predicted value of the first sensor for the existence probability of the target object.

[0048] Among them, the target object may be, for example, other traffic participants such as potential obstacles, pedestrians, and vehicles in the vehicle's surrounding environment; the target object may include one or multiple, and this application does not limit this. Schematically, the first sensor may be any one of the multiple sensors carried by the vehicle; the first data may be the data obtained after the first sensor detects the vehicle's surrounding environment; the initial probability of the target object in the vehicle's surrounding environment indicated by the first data may be, for example, the predicted value of the existence probability of the potential target object in the vehicle's surrounding environment determined according to the first data, and can be used as the initial value for determining the existence probability of the target object.

[0049] Schematically, the method for obtaining the initial probability of the target object indicated by the first data can be adjusted according to the type of the first sensor. In some embodiments, when the first sensor is a camera, the first data obtained by the camera detecting the vehicle surrounding environment can be processed by the YOLO algorithm to obtain the initial probability of the target object indicated by the first data. In some embodiments, when the first sensor is a lidar / millimeter-wave radar, the first data obtained by the lidar / millimeter-wave radar detecting the vehicle surrounding environment can be processed by the CFAR (Constant False Alarm Rate Detector) algorithm to obtain the initial probability of the target object indicated by the first data. The second sensor can be, for example, any one or more of the other sensors in the multiple sensors carried by the vehicle except the first sensor, and the second data can be, for example, the data obtained by the second sensor detecting the vehicle surrounding environment.

[0050] In some embodiments, the first sensor / second sensor can include, for example, any device or equipment such as a camera, a lidar, a millimeter-wave radar, etc. that can detect the vehicle surrounding environment, and the present application does not make any limitation in this regard.

[0051] Step S230, if the association between the second data and the target object fails, determine the first prediction probability of the target object according to the adjusted detection range of the second sensor, the first data, and the second data, where the first prediction probability is used to indicate the predicted value of the existence probability of the target object when the association between the second data and the target object fails for the second sensor.

[0052] When the association between the second data and the target object fails, the predicted value of the existence probability of the potential target object in the vehicle surrounding environment indicated by the first data for the second sensor should be 0 or a relatively small value close to 0. However, as described above, considering that the fluctuation of the detection range of the second sensor may cause the second sensor to fail to detect the potential target object in the vehicle surrounding environment indicated by the first data. In view of this, the embodiments of the present application can adjust the detection range of the second sensor according to the actual application situation (such as the driving scenario and / or vehicle type), and determine the first prediction probability of the potential target object in the vehicle surrounding environment indicated by the first data for the second sensor according to the adjusted detection range of the second sensor, the first data, and the second data, so as to ensure the accuracy and reliability of the first prediction probability, and further improve the accuracy and reliability when detecting the potential target object in the vehicle surrounding environment.

[0053] Considering that the false negative rate of the sensor for target objects at different positions within the detection range is different (the false negative rate can be used, for example, to indicate the probability of the sensor missing a target object present within the detection range), for example: the likelihood of missing a target object located in the edge region of the detection range is relatively high, while the likelihood of missing a target object located in the central region of the detection range is relatively low. Therefore, the false negative rate in the edge region of the sensor's detection range can be greater than that in the central region of the detection range. In view of this, based on the relative relationship between the position of the potential target object in the vehicle's surrounding environment indicated by the first data and the detection range of the adjusted second sensor, the reasonableness of the situation where the second sensor fails to detect the potential target object in the vehicle's surrounding environment indicated by the first data can be determined (for example, the reasonableness of the false negative situation of the second sensor in the edge region of the detection range is higher than that in the central region of the detection range), and then the first prediction probability of the second sensor for the potential target object in the vehicle's surrounding environment indicated by the first data can be determined.

[0054] In some embodiments, determining the first prediction probability of the target object based on the detection range of the adjusted second sensor, the first data, and the second data may include, for example: determining the position of the target object based on the first data; if the position of the target object indicated by the first data is within the detection range of the adjusted second sensor, determining the false negative rate of the second sensor based on the relative relationship between the position of the target object indicated by the first data and the detection range of the adjusted second sensor; and determining the first prediction probability based on the position of the target object indicated by the first data, the false negative rate of the second sensor, and the second data.

[0055] Among them, the method for determining the missed detection rate of the second sensor based on the relative relationship between the position of the target object indicated by the first data and the detection range of the adjusted second sensor can refer to the following criteria: The missed detection rate of the second sensor is positively correlated with the distance between the position of the target object indicated by the first data and the installation position of the second sensor, and / or the angle between the line connecting the position of the target object indicated by the first data and the installation position of the second sensor and the central normal line of the detection range of the adjusted second sensor. For example, the smaller the distance between the position of the target object indicated by the first data and the installation position of the second sensor, and / or the smaller the angle between the line connecting the position of the target object indicated by the first data and the installation position of the second sensor and the central normal line of the detection range of the adjusted second sensor, the closer the position of the target object indicated by the first data is to the central region of the detection range of the adjusted second sensor, and the corresponding first missed detection rate is smaller at this time; the larger the distance between the position of the target object indicated by the first data and the installation position of the second sensor, and / or the larger the angle between the line connecting the position of the target object indicated by the first data and the installation position of the second sensor and the central normal line of the detection range of the adjusted second sensor, the closer the position of the target object indicated by the first data is to the edge region of the detection range of the adjusted second sensor, and the corresponding second missed detection rate is higher at this time.

[0056] Through the above method, the present application can enable the second sensor to determine different missed detection rates according to the rationality of different missed detection situations. Furthermore, it ensures the accuracy and reliability of the first prediction probability of the second sensor for potential target objects in the vehicle surrounding environment indicated by the first data.

[0057] In some embodiments, when there are other objects in the detection range of the adjusted second sensor, which form an occlusion for potential target objects in the vehicle surrounding environment indicated by the first data, in order to further improve the accuracy and reliability of the first prediction probability of the second sensor for potential target objects in the vehicle surrounding environment indicated by the first data, determining the first prediction probability according to the position of potential target objects in the vehicle surrounding environment indicated by the first data, the missed detection rate of the second sensor, and the second data may include: determining the occlusion rate of the second sensor when detecting the target object according to the potential position in the vehicle surrounding environment indicated by the first data and the second data; determining the first prediction probability according to the occlusion rate, the missed detection rate of the second sensor, and the second data.

[0058] In some embodiments, if the position of the target object indicated by the first data is not within the detection range of the adjusted second sensor, it indicates that the target object indicated by the first data may be outside the detection range of the second sensor at this time, and the reference of the first prediction probability of the second sensor for the presence probability of the target object is relatively low. In view of this, in some embodiments, if the position of the target object is not within the detection range of the adjusted second sensor, the influence coefficient of the first prediction probability of the second sensor for the presence probability of the target object on the initial probability of the target object can be reduced. The influence coefficient is used to adjust the change range of the initial probability caused by the first prediction probability when the first prediction probability updates the initial probability.

[0059] Considering that the credibility of the predicted value of the presence probability of potential target objects in the vehicle's surrounding environment by the sensor may be affected by vehicle speed, weather, and sensor type, in view of this, in some embodiments, if the second data is successfully associated with the potential target object in the vehicle's surrounding environment indicated by the first data, the second prediction probability of the target object is determined according to the vehicle speed, driving scenario, and the second data. The second prediction probability is used to indicate the predicted value of the presence probability of the target object by the second sensor when the second data is successfully associated with the target object.

[0060] In some embodiments, the method for determining the second prediction probability of the target object according to the vehicle speed, driving scenario, and the second data may include, for example: obtaining a first confidence level related to the vehicle speed, a second confidence level related to the type of the sensor, a third confidence level related to the driving scenario, and the initial prediction probability of the second sensor for the target object; determining the second prediction probability of the second sensor for the potential target object in the vehicle's surrounding environment indicated by the first data according to the first confidence level, the second confidence level, the third confidence level, and the initial prediction probability. For example, the second prediction probability of the second sensor for the potential target object in the vehicle's surrounding environment indicated by the first data may be the product of the first confidence level, the second confidence level, the third confidence level, and the initial prediction probability.

[0061] Figure 3 It is a schematic diagram of the first confidence level curve provided by the embodiments of the present application.

[0062] As Figure 3 indicated, when the vehicle speed is within the interval [-50, 50], the first confidence level is 1 or close to 1. When the vehicle speed is outside the interval [-50, 50], the first confidence level is negatively correlated with the absolute value of the speed. It should be noted that Figure 3 the content shown in

[0063] In some embodiments, the first / second sensors in a vehicle may include, for example, a camera, a long-range radar (LRR), and a short-range radar (SRR). The driving scenarios of the vehicle may include, for example, clear weather, rain, fog, night, tunnels, and large lighting changes. The corresponding relationship between the third confidence levels of the camera, lidar, and millimeter-wave radar and the driving scenarios may refer to the content shown in the following table. Among them, the long-range radar may include lidar, and the short-range radar may include millimeter-wave radar.

[0064] LRR SRR Camera Sunny 1 1 1 Rain 0.98 0.98 0.8 Fog 1 1 0.7 Night 1 1 0.8 Tunnel 0.95 0.95 1 Large light change 1 1 0.7

[0065] Step S240: Update the initial probability according to the first prediction probability to determine whether the target object exists based on the updated initial probability.

[0066] In some embodiments, updating the initial probability according to the first prediction probability may be, for example, fusing the first prediction probability and the initial probability through the Dempster-Shafer Theory (DST).

[0067] In some embodiments, the method for determining whether the target object exists based on the updated initial probability may include: If the updated initial probability is less than the deletion threshold, determine that the target object corresponding to the initial probability less than the deletion threshold does not exist; if the updated initial probability is greater than the deletion threshold, continuously update the initial probability of the target object corresponding to the initial probability greater than the deletion threshold; if the updated initial probability is greater than the release threshold, determine that the target object corresponding to the initial probability greater than the release threshold exists. Among them, the deletion threshold is less than the release threshold. The deletion threshold is used to indicate the critical value for the non-existence of the target object. When the updated initial probability of the target object is less than the deletion threshold, it is considered that the target object does not exist, and the first sensor may have a detection error; the release threshold is used to indicate the critical value for the existence of the target object. When the updated initial probability of the target object is greater than the release threshold, it is considered that the target object exists; when the updated initial probability of the target object is greater than the deletion threshold but less than the release threshold, it is considered that the target object may or may not exist, and the initial probability of the target object can be continuously updated. The values of the deletion threshold and the release threshold can be adjusted according to the actual application situation.

[0068] Through the above method, the embodiments of the present application can determine whether there is a potential target object in the vehicle surrounding environment according to the updated initial probability of the target object, and by setting the release threshold to be greater than the deletion threshold, a certain buffer space is left for some doubtful target objects. By continuously updating the initial probability of the target object to further determine whether the target object exists, the situation of misjudgment can be reduced, and the accuracy and reliability of detection can be improved.

[0069] The technical solution disclosed by the embodiments of the present application fully considers the missed detection situation caused by the fluctuation of the detection range of the sensor. When the second data is associated with the target object unsuccessfully (that is, when the second sensor does not detect the target object), the detection range of the second sensor can be adjusted in a timely manner according to the driving scenario of the vehicle and / or the vehicle type, and the first predicted value of the existence probability of the target object by the second sensor can be determined according to the adjusted range, effectively avoiding the reduction of the existence probability of the target object caused by the missed detection situation caused by the fluctuation of the sensor detection range, and making the detection result of the target detection more reliable and accurate.

[0070] In some possible implementation manners, the present application further provides a control device for target detection in autonomous driving. Figure 4 is a schematic structural diagram of the control device for target detection in autonomous driving provided by the embodiments of the present application. Refer to Figure 4 , the control device for target detection in autonomous driving provided by the embodiments of the present application includes.

[0071] An adjustment module 410, configured to adjust the detection range of the second sensor according to the driving scenario and / or the vehicle type.

[0072] An acquisition module 420, configured to acquire the initial probability of the target object indicated by the first data and the second data. The first data corresponds to the first sensor after adjusting the detection range, and the second data corresponds to the second sensor after adjusting the detection range. The initial probability is used to indicate the predicted value of the existence probability of the target object by the first sensor.

[0073] A determination module 430, configured to, if the second data is associated with the target object unsuccessfully, determine the first predicted probability of the target object according to the adjusted detection range of the second sensor, the first data, and the second data. The first predicted probability is used to indicate the predicted value of the existence probability of the target object by the second sensor when the second data is associated with the target object unsuccessfully.

[0074] An update module 440, configured to update the initial probability according to the first predicted probability, so as to determine whether the target object exists according to the updated initial probability.

[0075] In some embodiments, the adjustment module 410 is configured to, if the driving scenario is an overtaking scenario, reduce the detection ranges of the first sensor and the second sensor.

[0076] In some embodiments, the adjustment module 410 is configured to reduce the detection ranges of the first sensor and the second sensor if the vehicle type is a bus or a truck.

[0077] In some embodiments, the determination module 430 is configured to determine the position of the target object according to the first data; if the position is within the detection range of the adjusted second sensor, determine the missed detection rate of the second sensor according to the relative relationship between the position and the detection range of the adjusted second sensor; and determine the first prediction probability according to the position, the missed detection rate, and the second data.

[0078] In some embodiments, the determination module 430 is configured to reduce the influence coefficient of the first prediction probability on the initial probability if the position is not within the detection range of the adjusted second sensor.

[0079] In some embodiments, the determination module 430 is configured to determine the occlusion rate of the second sensor when detecting the target object according to the position and the second data; and determine the first prediction probability according to the occlusion rate, the missed detection rate, and the second data.

[0080] In some embodiments, the determination module 430 is further configured to determine the second prediction probability of the target object according to the speed of the vehicle, the driving scenario, and the second data if the second data is successfully associated with the target object, where the second prediction probability is used to indicate the predicted value of the existence probability of the target object when the second data is successfully associated with the target object by the second sensor.

[0081] In some embodiments, the update module 440 is configured to determine that the target object corresponding to the initial probability less than the deletion threshold does not exist if the updated initial probability is less than the deletion threshold; continuously update the initial probability of the target object corresponding to the initial probability greater than the deletion threshold if the updated initial probability is greater than the deletion threshold; and determine that the target object corresponding to the initial probability greater than the release threshold exists if the updated initial probability is greater than the release threshold, where the deletion threshold is less than the release threshold, the deletion threshold is used to indicate the critical value of the non - existence of the target object, and the release threshold is used to indicate the critical value of the existence of the target object.

[0082] It should be understood that when the above - provided device implements its functions, only the division of the above - mentioned functional modules is used for illustration. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the control device for target detection in autonomous driving provided in the above embodiments and the method embodiments for target detection in autonomous driving belong to the same concept, and the specific implementation process can be found in the method embodiments for target detection in autonomous driving.

[0083] In some other possible embodiments, the present application further provides an electronic device for target detection in autonomous driving. Figure 5 is a schematic structural diagram of the electronic device for target detection in autonomous driving provided by the embodiments of the present application. Refer to Figure 5 , the electronic device for target detection in autonomous driving provided by the embodiments of the present application includes:

[0084] A memory 510, on which at least one program instruction for target detection in autonomous driving is stored.

[0085] A processor 520, when the above program instruction is executed by the processor 520, enables the electronic device to implement the methods and steps of its multiple embodiments described above in conjunction with Figure 2 . Depending on the implementation, the processor 520 may be a CPU (central processing unit), a GPU (graphics processing unit), or one or more types of other general-purpose and / or special-purpose processors, including but not limited to a DSP (digital signal processor), an ASIC (application specific integrated circuit), an FPGA (field-programmable gate array), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., and the number thereof can be determined according to actual needs.

[0086] In still some other possible embodiments, the present application further provides a computer program (product), the computer program (product) includes computer programs / instructions, and when the computer programs / instructions are executed by a processor, enables the computer to implement the methods and steps of its multiple embodiments described above in conjunction with Figure 2 .

[0087] In still some other possible embodiments, the present application further provides a computer-readable storage medium, on which program instructions for target detection in autonomous driving are stored. When the program instructions are executed by one or more processors, the methods described above in conjunction with Figure 2The steps of the described method and its various embodiments. The computer-readable storage medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0088] It should be noted that the electronic device in this application may also be referred to as a display device. In addition, the information, data (including but not limited to image data, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions. For example, the grayscale values involved in this application are obtained under full authorization.

[0089] The term "and / or" in the embodiments of this application is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone.

[0090] The above is only for the convenience of those skilled in the art to understand the technical solution of this application, and is not intended to limit this application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included in the protection scope of this application.

Claims

1. A method for target detection in autonomous driving, characterized in that: The method is used to detect a target object in a vehicle surrounding environment, the vehicle being equipped with at least a first sensor and a second sensor, and the method comprises: Adjusting the detection ranges of the first sensor and the second sensor according to the driving scene and / or the vehicle type; Acquire an initial probability of the target object indicated by first data and second data, wherein the first data corresponds to the first sensor after the detection range is adjusted, and the second data corresponds to the second sensor after the detection range is adjusted, and the initial probability is used to indicate a predicted value of the first sensor for the existence probability of the target object; If the second data fails to be associated with the target object, determining a first predicted probability of the target object according to the adjusted detection range of the second sensor, the first data, and the second data, the first predicted probability being used to indicate a predicted value of the existence probability of the target object by the second sensor when the second data fails to be associated with the target object; The initial probability is updated according to the first predicted probability to determine whether the target object exists according to the updated initial probability.

2. The method according to claim 1, characterized in that The adjusting the detection ranges of the first sensor and the second sensor according to the driving scene and / or the vehicle type includes: If the driving scene is an overtaking scene, the detection ranges of the first sensor and the second sensor are reduced.

3. The method according to claim 1, characterized in that The adjusting the detection ranges of the first sensor and the second sensor according to the driving scene and / or the vehicle type includes: If the vehicle type is a bus or a truck, the detection ranges of the first sensor and the second sensor are reduced.

4. The method according to claim 1, characterized in that: The determining a first predicted probability of the target object according to the adjusted detection range of the second sensor, the first data, and the second data includes: determining the position of the target object according to the first data; If the position is within the adjusted detection range of the second sensor, determining the missed detection rate of the second sensor according to a relative relationship between the position and the adjusted detection range of the second sensor; The first predicted probability is determined according to the position, the missed detection rate and the second data.

5. The method according to claim 4, characterized in that If the position is not within the adjusted detection range of the second sensor, the influence coefficient of the first predicted probability on the initial probability is reduced.

6. The method according to claim 4, characterized in that The determining the first prediction probability according to the position, the missed detection rate and the second data includes: determining, according to the position and the second data, an occlusion rate of the second sensor when detecting the target object; The first prediction probability is determined according to the occlusion rate, the missed detection rate and the second data.

7. The method according to claim 1, characterized in that The method further comprises: If the second data is successfully associated with the target object, a second predicted probability of the target object is determined based on the vehicle speed, the driving scene and the second data, and the second predicted probability is used to indicate the predicted value of the existence probability of the target object by the second sensor when the second data is successfully associated with the target object.

8. The method according to claim 1, characterized in that: Determining whether the target object exists according to the updated initial probability includes: If the updated initial probability is less than the deletion threshold, determining that the target object corresponding to the initial probability less than the deletion threshold does not exist; If the updated initial probability is greater than the deletion threshold, continuously updating the initial probability of the target object corresponding to the initial probability greater than the deletion threshold; If the initial probability after updating is greater than the publishing threshold, it is determined that the target object corresponding to the initial probability greater than the publishing threshold exists, wherein the deletion threshold is less than the publishing threshold, the deletion threshold is used to indicate the critical value of the non-existence of the target object, and the publishing threshold is used to indicate the critical value of the existence of the target object.

9. An electronic device, characterized in that: include: A memory having stored thereon program instructions for target detection in autonomous driving; as well as The processor, when the program instructions are executed by the processor, enables the vehicle to implement the method described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: Program instructions for target detection in autonomous driving are stored thereon, and when the program instructions are executed by one or more processors, the vehicle implements the method described in any one of claims 1-8.

Images