Obstacle identification method and system for vehicle adaptive cruise

Abstract

The invention relates to an obstacle recognition method and system for vehicle adaptive cruise. The method includes acquiring a road image of a vehicle based on the vehicle being in an adaptive cruise state. Determining whether the light of the environment where the vehicle is located is abnormal based on the road image so as to eliminate the influence of the environment on the recognition precision; in response to the fact that the light of the environment where the vehicle is located is determined to be abnormal, the confidence coefficient corresponding to the type of at least one to-be-recognized object is recognized from the road image; and determining whether the to-be-identified object is a to-be-processed obstacle or not based on the confidence coefficient corresponding to the type of the to-be-identified object and the at least two state parameters. Whether the to-be-recognized object is the to-be-processed obstacle needing to be filtered or not is judged more accurately by combining the confidence coefficient corresponding to the type of the to-be-recognized object and the various state parameters, the mistaken braking phenomenon of the vehicle in the high-speed driving state is effectively reduced, and more stable and comfortable self-adaptive cruise function experience is brought to a driver.

Description

Technical Field

[0001] The embodiments in this specification belong to the field of vehicle intelligent driving control technology, and specifically relate to an obstacle recognition method and system for vehicle adaptive cruise control. Background Technology

[0002] Advanced Driver Assistance Systems (ADAS) have been widely adopted and promoted as a core technology of modern automotive intelligence. Especially during high-speed driving, the adaptive cruise control function of ADAS can automatically adjust the vehicle speed to maintain a safe distance from the vehicle in front based on the driver's set fixed speed and distance.

[0003] When using the adaptive cruise control function of an advanced driver assistance system, the vehicle is traveling at high speed, which can easily lead to accidental braking and thus affect the driver's experience with the adaptive cruise control function. Summary of the Invention

[0004] The embodiments of this disclosure provide an obstacle recognition method and system for vehicle adaptive cruise control.

[0005] In a first aspect of this disclosure, an obstacle recognition method for adaptive cruise control of a vehicle is provided. The method includes acquiring a road image of the vehicle based on the vehicle being in adaptive cruise control mode. The method further includes determining whether the lighting conditions in the vehicle's environment are abnormal based on the road image. In response to determining that the lighting conditions in the vehicle's environment are abnormal, the method further includes identifying a confidence level corresponding to the type of at least one object to be identified from the road image. Furthermore, the method includes determining whether the object to be identified is an obstacle to be processed based on the confidence level corresponding to the type of the object to be identified and at least two state parameters.

[0006] In a second aspect of this disclosure, an obstacle recognition system for adaptive cruise control of a vehicle is provided. The system includes an image acquisition module configured to acquire a road image of the vehicle based on the vehicle being in adaptive cruise control mode. The system also includes a light determination module configured to determine whether the lighting in the vehicle's environment is abnormal based on the road image. The system further includes an object recognition module configured to, in response to determining abnormal lighting in the vehicle's environment, identify a confidence level corresponding to the type of at least one object to be recognized from the road image. Furthermore, the system includes an obstacle determination module configured to determine whether the object to be recognized is an obstacle to be processed based on the confidence level corresponding to the type of the object to be recognized and at least two state parameters.

[0007] In a third aspect of this disclosure, a computer program product is provided, comprising a computer program that is executed by a processor to implement the method according to the first aspect.

[0008] In a fourth aspect of this disclosure, a machine-readable storage medium is provided. The machine-readable storage medium stores machine-executable instructions, which are executed by a processor to implement the method provided according to a first aspect of this disclosure.

[0009] It should be understood that the description in the Summary of the Invention section is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0010] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein: Figure 1 A schematic diagram of an example environment in which some embodiments of this disclosure may be implemented is shown; Figure 2 A flowchart illustrating an obstacle recognition method for vehicle adaptive cruise control according to some embodiments of this disclosure is shown. Figure 3 This illustration shows a schematic diagram of the correspondence between some embodiments of the present disclosure for determining whether an object to be identified is an obstacle to be processed; Figure 4 A block diagram of an obstacle recognition system for vehicle adaptive cruise control, according to some embodiments of this disclosure, is shown; and Figure 5 A block diagram of an electronic device that can implement several embodiments of the present disclosure is shown. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0012] The terms “comprising” and “having”, and any variations thereof, in this specification, claims, and the foregoing drawings are intended to cover a non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. Depending on the context, the word “if” as it applies herein may be interpreted as “when”, “when”, “in response to determination”, or “in response to detection”.

[0013] As mentioned above, when using the adaptive cruise control function of an advanced driver assistance system, because the vehicle is traveling at high speed, it is easy to misidentify objects that are not in the lane into the lane when visibility is poor, or to misidentify non-vehicle objects in the lane as target objects. This can cause the adaptive cruise control function to brake unnecessarily, which not only poses certain safety hazards to the vehicle at high speed, but also affects the driver's experience with the adaptive cruise control function.

[0014] Therefore, embodiments of this disclosure propose an obstacle recognition method for vehicle adaptive cruise control. The method includes acquiring a road image of the vehicle based on the vehicle being in adaptive cruise control mode. The method further includes determining whether the lighting conditions in the vehicle's environment are abnormal based on the road image. In response to determining that the lighting conditions in the vehicle's environment are abnormal, the method further includes identifying a confidence level corresponding to the type of at least one object to be identified from the road image. Furthermore, the method includes determining whether the object to be identified is an obstacle to be processed based on the confidence level corresponding to the type of the object to be identified and at least two state parameters.

[0015] In this way, it is possible to make a preliminary judgment on whether the ambient light is abnormal based on road images, so as to eliminate the influence of the environment on the recognition accuracy. When it is determined that the ambient light is abnormal, it can also identify the confidence level corresponding to the type of the object to be identified, and combine multiple state parameters to more accurately determine whether the object to be identified is an obstacle to be filtered. This can not only effectively reduce the phenomenon of false braking of the vehicle at high speed, but also bring the driver a more stable and comfortable adaptive cruise control experience.

[0016] Figure 1 Schematic diagrams are shown illustrating example environments in which some embodiments of this disclosure can be implemented. For example... Figure 1As shown, the example environment 100 may include an onboard sensor 101, which is used to acquire a road image in front of the vehicle, the vehicle's current speed, and state parameters of one or more objects to be identified when the vehicle is in adaptive cruise control mode. Here, the onboard sensor 101 includes at least a front-facing camera, a wheel speed sensor, and radar (divided into lidar and millimeter-wave radar). The front-facing camera is used to acquire a road image in front of the vehicle at the current acquisition time, the wheel speed sensor is used to acquire the vehicle's current speed at the current acquisition time, and the radar is used to acquire state parameters of one or more objects to be identified in front of the vehicle at the current acquisition time. The state parameters of each object to be identified include one or more of the following: current speed, relative speed to the vehicle, longitudinal distance, lateral distance, distance standard deviation, and heading angle. It is understandable that the object to be identified can be understood as an object that may be identified as an obstacle (such as a vehicle, shadow, or puddles) located directly in front of the vehicle's lane. Of course, it can also be understood as an object that may be identified as an obstacle (such as a vehicle, shadow, or puddles) located directly in front of the vehicle's lane and one or more objects that may be identified as obstacles (such as vehicles, shadows, or puddles) located in front of the vehicle in different lanes. For example, all objects in the acquired road image excluding calibrated reference objects such as trees, houses, and railings can be used as the object to be identified. The current speed, relative speed with the vehicle, and distance standard deviation of each object to be identified can be obtained by millimeter-wave radar, and the state parameters such as longitudinal distance, lateral distance, and azimuth angle between each object to be identified and the vehicle can be obtained by lidar, and it is not limited to these.

[0017] Example environment 100 may further include a control terminal 102, which can be understood as an intelligent driving domain controller in the art, such as a controller for a vehicle's advanced driver assistance system. It acquires road images and the vehicle's current speed, as well as state parameters of one or more objects to be identified, from the onboard sensor 101 by establishing a communication connection with the onboard sensor 101. Here, when the vehicle is traveling at high speed, the control terminal 102 can control the vehicle to be in adaptive cruise control mode based on the driver's selected driver assistance activation command. When it determines that the object directly in front of the vehicle's lane is not an obstacle to be filtered, it generates a control command based on the driver's set vehicle speed, distance, and the state parameters of the object directly in front of the vehicle's lane, and sends it to the vehicle actuator 103, so that the vehicle actuator 103 can control the vehicle to maintain a distance from the object directly in front of it in the lane. Of course, in some embodiments of this disclosure, the control terminal 102 can also establish a communication connection with the vehicle sensor 101 through the vehicle's perception domain controller, so that the perception domain controller can analyze and process the data collected by the radar in the vehicle sensor 101 to obtain the status parameters of one or more objects to be identified in front of the vehicle, and the perception domain controller can feed back the status parameters of one or more objects to be identified in front of the vehicle to the control terminal 102, but this will not be elaborated further here.

[0018] Furthermore, after acquiring the road image in front of the vehicle, the vehicle's current speed, and the state parameters of one or more objects to be identified, the control terminal 102 can also acquire the road image of the vehicle based on the vehicle being in adaptive cruise control, to determine whether the lighting in the vehicle's environment is abnormal. Then, in response to determining that the lighting in the vehicle's environment is abnormal, it can identify the confidence level corresponding to the type of at least one object to be identified from the road image, and determine whether the object to be identified is an obstacle to be processed based on the confidence level corresponding to the type of the object to be identified and at least two state parameters. Here, after determining that the object to be identified is an obstacle to be processed, the object corresponding to the obstacle to be processed can be eliminated from all objects in front of the vehicle; that is, the control terminal 102 does not consider the impact of the obstacle to be processed on the vehicle's current driving state when controlling the vehicle in adaptive cruise control. Of course, in some embodiments of this disclosure, the control terminal can also mark the object corresponding to the obstacle to be processed after it is identified as an obstacle to be processed, so that the impact of the marked object on the vehicle's current driving state is no longer considered. For example, when an obstacle to be processed is directly in front of the vehicle's lane, the obstacle can be filtered out. When it is determined that the object directly in front of the obstacle to be processed is not an obstacle to be processed (e.g., a moving vehicle), a control command is generated based on the vehicle speed, distance, and state parameters of the object directly in front of the obstacle to be processed set by the driver and sent to the vehicle actuator 103, so that the vehicle actuator 103 controls the vehicle to maintain a distance from the object directly in front of the obstacle to be processed.

[0019] Example environment 100 may further include a vehicle actuator 103, which obtains control commands issued by the control terminal 102 through a communication connection, and controls the vehicle to maintain a distance from objects directly in front of it in the lane based on the control commands. Here, the vehicle actuator 103 is divided into an engine controller or motor controller for controlling vehicle acceleration, and a vehicle electronic stability controller or brake controller for controlling vehicle deceleration. When the control command is determined to be to reduce the vehicle speed, the vehicle electronic stability controller or brake controller controls the vehicle to reduce to a specified speed; when the control command is determined to be to increase the vehicle speed, the engine controller or motor controller controls the vehicle to increase to a specified speed.

[0020] In this way, it is possible to make a preliminary judgment on whether the ambient light is abnormal based on road images, so as to eliminate the influence of the environment on the recognition accuracy. When it is determined that the ambient light is abnormal, it can also identify the confidence level corresponding to the type of the object to be identified, and combine multiple state parameters to more accurately determine whether the object to be identified is an obstacle to be filtered. This can not only effectively reduce the phenomenon of false braking of the vehicle at high speed, but also bring the driver a more stable and comfortable adaptive cruise control experience.

[0021] It should be understood that the architecture and functionality in example environment 100 are described for illustrative purposes only and do not imply any limitation on the scope of this disclosure. Embodiments of this disclosure can also be applied to other environments with different architectures and / or functionalities.

[0022] Figure 2 A flowchart of an obstacle recognition method for vehicle adaptive cruise control according to some embodiments of the present disclosure is shown. Method 200 may, for example, be derived from... Figure 1 The control terminal in the example environment shown executes the commands. Figure 2 As shown in block 202, method 200 can acquire a road image of the vehicle based on the vehicle being in adaptive cruise control mode. In some implementations, the control terminal can determine whether the vehicle is in a high-speed driving state based on the vehicle's current position and current speed when the vehicle is in a normal driving state. If it is determined that the vehicle is in a high-speed driving state, the control terminal can control the vehicle to enter adaptive cruise control mode based on the driver's selected driving assistance activation command on the vehicle, and through the above... Figure 1 The example environment shown uses onboard sensors to acquire images of the road in front of the vehicle at the current acquisition time. Here, the road image may include multiple lane lines, one or more objects within one or more lane lines, trees, houses, and railings, etc.

[0023] Understandably, when the vehicle is currently on a highway and its current speed is not less than a preset speed threshold, the control terminal can determine that the vehicle is traveling at high speed, thus fulfilling the initial conditions for adaptive cruise control. When the vehicle is not currently on a highway, or its current speed is less than the preset speed threshold, the control terminal can determine that the vehicle is not traveling at high speed, thus failing to trigger adaptive cruise control.

[0024] In box 204, method 200 can determine whether the lighting in the vehicle's environment is abnormal based on the road image. Understandably, when it is determined that the lighting in the vehicle's environment is abnormal (e.g., dim lighting caused by nighttime), it indicates that the recognition accuracy for objects directly in front of the vehicle in its lane is poor, which easily leads to a higher risk of false braking. When it is determined that the lighting in the vehicle's environment is normal, it indicates that there is a certain recognition accuracy for objects directly in front of the vehicle in its lane, but it still cannot guarantee that there will be no false braking, and it is still necessary to further determine whether the object in front of the vehicle is an obstacle to be processed.

[0025] In some implementations, when the control terminal determines whether the lighting in the vehicle's environment is abnormal based on a road image, it can perform color space conversion processing on the road image and statistically determine the percentage of pixels corresponding to the low-light illumination range from the color space-converted road image. In one example, after acquiring a road image of the area in front of the vehicle at the current acquisition time, captured by a front-facing camera in the vehicle's onboard sensors, distortion correction processing can be performed on the road image based on the camera's intrinsic parameters and preset distortion parameters to avoid errors in subsequent data calculations caused by image distortion. Furthermore, color space conversion processing can be performed on the distortion-corrected road image to convert it from the RGB color space to the YUV color space, thereby improving the accuracy of image brightness recognition.

[0026] Next, the proportion of pixels corresponding to the low-light intensity range can be statistically determined from the road image after color space conversion. For example, all pixels whose Y channel values ​​are within a preset range can be taken as all pixels in the low-light intensity range, and the ratio between the number of all pixels in the low-light intensity range and the total number of pixels in the road image can be calculated. The ratio result is then determined as the proportion of pixels corresponding to the low-light intensity range.

[0027] The control terminal can then determine whether the percentage of pixels corresponding to the low-light illumination range is less than a preset percentage threshold. It can be understood that when the percentage of pixels corresponding to the low-light illumination range is less than the preset percentage threshold, it indicates that the ambient light around the vehicle is normal; when the percentage of pixels corresponding to the low-light illumination range is not less than the preset percentage threshold, it indicates that the ambient light around the vehicle is abnormal.

[0028] It should be noted that when the ambient light is normal, it indicates that the control terminal can accurately determine whether an object in front of the vehicle is an obstacle when the vehicle is in adaptive cruise control mode, thus ensuring the normal execution of adaptive cruise control.

[0029] Of course, some embodiments of this disclosure may also employ other technical means well known in the art to analyze and process road images in order to determine whether the lighting in the environment where the vehicle is located is normal, but these will not be elaborated upon here.

[0030] In block 206, method 200 may, in response to determining an anomaly in the lighting of the vehicle's environment, identify a confidence level corresponding to the type of at least one object to be identified from the road image. In some implementations, the control terminal may, in response to determining an anomaly in the lighting of the vehicle's environment, input the road image into a convolutional neural network to predict the confidence level corresponding to the type of each object in the road image. Here, the type of each object may be a car, a motorcycle, or a truck, and the confidence level corresponding to each object type can be understood to include the confidence level corresponding to a car, a motorcycle, or a truck. It is understood that a higher confidence level corresponding to an object type indicates a higher probability that the corresponding object is an obstacle to be processed; a lower confidence level indicates a lower probability that the corresponding object is an obstacle to be processed.

[0031] Understandably, the structure of a convolutional neural network (CNN) is a well-known network architecture in this field. It can process input road images and predict the confidence scores of multiple objects in the road images, excluding identifiable reference objects such as trees, houses, and railings. When training the CNN, the training set includes sample road images of multiple vehicles in normally lit environments, the outlines of multiple objects marked in each sample road image, and the confidence scores corresponding to the types of each object. The loss function can be set to calculate the mean squared error between the predicted confidence scores of each object type and the corresponding confidence scores. Based on the input sample road images of vehicles in abnormally lit environments, the system predicts multiple objects in the corresponding sample road images and the confidence scores corresponding to the types of each object. Then, the loss function calculates the total loss based on the confidence scores of each object type and the corresponding confidence scores. An optimizer is then used to update the weights of the CNN based on the total loss to minimize the total loss until the maximum number of training epochs is reached.

[0032] After determining the confidence level corresponding to the type of each object in the road image, the control terminal can filter one or more objects as objects to be identified based on the lane in which the vehicle is located. In one example, the object that is in the vehicle's lane and has the closest longitudinal distance to the vehicle can be identified as an object to be identified; alternatively, one or more objects in the vehicle's lane and one or more objects in adjacent lanes can be identified as multiple objects to be identified; or, all objects in the road image can be identified as multiple objects to be identified, and each object can be further determined to be an obstacle. Here, when there are multiple objects to be identified, the control terminal can also eliminate potential risks posed by objects in adjacent lanes to the vehicle's lane when it is in adaptive cruise control by determining whether each object is an obstacle to be processed.

[0033] Of course, some embodiments of this disclosure may also identify one or more objects in the lane where the vehicle is located, and one or more objects in adjacent lanes that have a tendency to change lanes as multiple objects to be identified. Here, one or more objects in adjacent lanes that have a tendency to change lanes can be identified by identifying whether the lateral distance between each object in adjacent lanes and the vehicle gradually decreases. For example, when it is identified that the historical lateral distance between an object in adjacent lanes and the vehicle gradually decreases, it indicates that the object has a tendency to change lanes and is moving closer to the vehicle. Thus, the object can be identified as an object with a tendency to change lanes, i.e., an object to be identified. The historical lateral distance between the object and the vehicle can be the lateral distance collected by the radar in the vehicle sensor at multiple consecutive historical collection times.

[0034] In block 208, method 200 can determine whether an object to be identified is an obstacle to be processed based on the confidence level corresponding to the type of the object to be identified and at least two state parameters. In some implementations, when the control terminal determines whether an object to be identified is an obstacle to be processed based on the confidence level corresponding to the type of the object to be identified and at least two state parameters, it can determine whether the confidence level corresponding to the type of the object to be identified is less than a preset confidence level threshold.

[0035] Understandably, when the confidence level corresponding to the type of the object to be identified is not less than the preset confidence threshold, it indicates that it is still impossible to determine whether the object to be identified is an obstacle to be processed; when the confidence level corresponding to the type of the object to be identified is less than the preset confidence threshold, it indicates that it can be determined that the object to be identified is not an obstacle to be processed, that is, there is no need to filter the object to be identified. Furthermore, when the object to be identified is an object within the lane where the vehicle is located and is the closest to the vehicle in longitudinal distance, a control command can be generated based on the state parameters of the object to be identified, the vehicle speed set by the driver, and the time distance, and sent to the vehicle actuator. The vehicle actuator then controls the vehicle to maintain the distance between itself and the object to be identified.

[0036] Subsequently, in response to determining that the confidence level corresponding to the type of the object to be identified is not less than a preset confidence threshold, the control terminal may further determine whether the object to be identified is an obstacle to be processed based on at least two state parameters. Here, the at least two state parameters include at least two of the following: the current speed of the object to be identified, the current speed of the vehicle, the relative speed between the vehicle and the object to be identified, the longitudinal distance, and the standard deviation of the distance.

[0037] In some implementations, when the control terminal determines whether the object to be identified is an obstacle to be processed based on at least two state parameters, it can determine whether the longitudinal distance is less than a distance threshold corresponding to the type of the object to be identified. Here, the distance threshold corresponding to the type of the object to be identified can be understood as follows: when the type of the object to be identified is a car, the corresponding distance threshold can be set to 150; when the type of the object to be identified is a motorcycle or a truck, the corresponding distance threshold can be set to 130.

[0038] Understandably, when the longitudinal distance is less than the distance threshold corresponding to the type of the object to be identified, it indicates that it is not yet possible to determine whether the object to be identified is an obstacle to be processed, and it is necessary to further combine other state parameters to make a judgment. When the longitudinal distance is not less than the distance threshold corresponding to the type of the object to be identified, it indicates that the object to be identified is an obstacle to be processed, and then the obstacle to be processed can be filtered out. For example, the obstacle to be processed can be removed from all objects in front of the vehicle so that the impact of the obstacle to be processed on the current driving state of the vehicle is not considered when the vehicle is in adaptive cruise control.

[0039] Subsequently, the control terminal can determine whether the current speed of the object to be identified is less than a preset vehicle speed threshold in response to determining that the longitudinal distance is less than a distance threshold corresponding to the type of the object to be identified. Here, the preset vehicle speed threshold can be set to 100, but is not limited to this. It is understood that when the current speed of the object to be identified is less than the preset vehicle speed threshold, it indicates that it is not yet possible to determine whether the object to be identified is an obstacle to be processed, and it is necessary to further combine other state parameters to make a judgment; when the current speed of the object to be identified is not less than the preset vehicle speed threshold, it indicates that it can be determined that the object to be identified is an obstacle to be processed, and then the obstacle to be processed can be filtered, for example, the obstacle to be processed can be removed from all objects in front of the vehicle, so that when the vehicle is in adaptive cruise control mode, the impact of the obstacle to be processed on the current driving state of the vehicle is not considered.

[0040] Subsequently, the control terminal can respond to the determination that the current speed of the object to be identified is less than a preset vehicle speed threshold, and determine whether the object to be identified is an obstacle to be processed based on the current speed of the vehicle, the type of the object to be identified, the current speed of the object to be identified, the relative speed between the vehicle and the object to be identified, the longitudinal distance, and the distance standard deviation.

[0041] In some implementations, when the control terminal determines whether an object to be identified is an obstacle based on the vehicle's current speed, the type of the object to be identified, the current speed of the object to be identified, the relative speed between the vehicle and the object to be identified, the longitudinal distance, and the distance standard deviation, it can determine a distance standard deviation threshold expression for the object to be identified based on the vehicle's current speed and the type of the object to be identified. Based on the distance standard deviation and the distance standard deviation threshold expression, it can then determine a target distance standard deviation threshold. In one example, a distance standard deviation threshold expression corresponding to the vehicle's current speed and the type of the object to be identified can be retrieved from a preset speed-type-distance standard deviation expression correspondence. The distance standard deviation between the object to be identified and the vehicle is then substituted into this distance standard deviation threshold expression to obtain the target distance standard deviation threshold. This preset speed-type-distance standard deviation expression correspondence can be derived through manual experience combined with historical data analysis. For example, when the vehicle's current speed is 115 and the type of the object to be identified is a sedan, the corresponding distance standard deviation threshold expression can be expressed as 110 - min(distance standard deviation, 30); when the vehicle's current speed is 115 and the type of the object to be identified is a motorcycle, the corresponding distance standard deviation threshold expression can be expressed as 110 - min(2*distance standard deviation, 30); when the vehicle's current speed is 115 and the type of the object to be identified is a truck, the corresponding distance standard deviation threshold expression can be expressed as 110 - min(2*distance standard deviation, 30), and it is not limited to these expressions.

[0042] Subsequently, the control terminal can determine the actual distance standard deviation based on the longitudinal distance and determine whether the actual distance standard deviation is less than the target distance standard deviation threshold. In one example, the difference between 100 and the longitudinal distance can be determined as the actual distance standard deviation, and it can be determined whether this actual distance standard deviation is less than the target distance standard deviation threshold. It is understood that when the actual distance standard deviation is less than the target distance standard deviation threshold, it can be determined that the object to be identified is not an obstacle to be processed; when the actual distance standard deviation is not less than the target distance standard deviation threshold, it indicates that it is still impossible to determine whether the object to be identified is an obstacle to be processed, and further judgment needs to be made in conjunction with other state parameters.

[0043] Subsequently, the control terminal, in response to determining that the actual distance standard deviation is not less than the target distance standard deviation threshold, determines the target speed standard deviation threshold for the object to be identified based on the vehicle's current speed and the type of the object to be identified. In one example, a speed standard deviation threshold corresponding to the vehicle's current speed and the type of the object to be identified can be retrieved from a preset speed-type-speed standard deviation threshold correspondence, and this threshold is determined as the target speed standard deviation threshold for the object to be identified. This preset speed-type-speed standard deviation threshold correspondence can be derived through manual experience combined with historical data analysis. For example, when the vehicle's current speed is 115 and the type of the object to be identified is a sedan, the corresponding speed standard deviation threshold can be 60; when the vehicle's current speed is 115 and the type of the object to be identified is a motorcycle, the corresponding speed standard deviation threshold can be 50; when the vehicle's current speed is 115 and the type of the object to be identified is a truck, the corresponding speed standard deviation threshold can be 60, and it is not limited to these values.

[0044] Subsequently, the control terminal can determine the actual speed standard deviation of the object to be identified based on the vehicle's current speed, the object's current speed, relative speed, and longitudinal distance, and determine whether the actual speed standard deviation is less than a target speed standard deviation threshold. In one example, the vehicle's current speed, the object's current speed, relative speed, and longitudinal distance can be input into a deep learning model to predict the actual speed standard deviation of the object through an intelligent algorithm. Here, the deep learning model structure can be a well-known model structure in the art, capable of performing correlation analysis on the input vehicle's current speed, the object's current speed, relative speed, and longitudinal distance to predict the actual speed standard deviation of the object. When training a deep learning model, the training set includes multiple sets of sample state parameters (each set of sample state parameters includes the current speed of the vehicle, the current speed of the object to be identified, the relative speed of the sample, and the longitudinal distance of the sample), and the speed standard deviation label of the object to be identified corresponding to each set of sample state parameters. The loss function can be set to calculate the mean square error between the predicted speed standard deviation of the object to be identified corresponding to each set of sample state parameters and the corresponding speed standard deviation label. The speed standard deviation of the object to be identified is predicted based on the input sample state parameters. Then, the total loss is calculated by the loss function based on the speed standard deviation and the corresponding speed standard deviation label of each set of sample state parameters. The optimizer is then used to update the weights of the deep learning model based on the total loss to minimize the total loss until the maximum number of training epochs is reached.

[0045] Of course, some embodiments of this disclosure can also use a preset correspondence to query the speed standard deviation corresponding to the current speed of the vehicle, the current speed of the object to be identified, the relative speed and the longitudinal distance, and determine the speed standard deviation as the actual speed standard deviation of the object to be identified, but are not limited to this.

[0046] Understandably, when the actual speed standard deviation is not less than the target speed standard deviation threshold, the object to be identified can be determined as the obstacle to be processed. Then, the control terminal can filter the obstacle to be processed, for example, by removing the obstacle from all objects in front of the vehicle, so that the impact of the obstacle to be processed on the current driving state of the vehicle is not considered when the vehicle is in adaptive cruise control mode.

[0047] When the actual speed standard deviation is less than the target speed standard deviation threshold, it can be determined that the object to be identified is not an obstacle to be processed.

[0048] It is also understandable that after determining that the object to be identified is not an obstacle to be processed, the control terminal does not need to filter the object to be identified. Furthermore, when it is identified that the object to be identified is the object within the lane where the vehicle is located and is the closest to the vehicle in longitudinal distance, the control terminal can generate control commands based on the state parameters of the object to be identified, the vehicle speed set by the driver, and the time distance and send them to the vehicle actuator in the example environment mentioned above. The vehicle actuator then controls the vehicle to maintain the distance between itself and the object to be identified.

[0049] Please see Figure 3 The diagram illustrates a correspondence for determining whether an object to be identified is an obstacle to be processed, according to some embodiments of this disclosure. Figure 3 As shown, the correspondence 300 for determining whether the object to be identified is an obstacle to be processed includes a distance standard deviation expression and a speed standard deviation threshold corresponding to the current speed of the vehicle and the type of the object to be identified. For example, when the current speed of the vehicle is 115 and the type of the object to be identified is a sedan, the corresponding distance standard deviation threshold expression can be expressed as 110 - min(distance standard deviation, 30), and the corresponding speed standard deviation threshold can be 60; when the current speed of the vehicle is 115 and the type of the object to be identified is a motorcycle, the corresponding distance standard deviation threshold expression can be expressed as 110 - min(2*distance standard deviation, 30), and the corresponding speed standard deviation threshold can be 50; when the current speed of the vehicle is 115 and the type of the object to be identified is a truck, the corresponding distance standard deviation threshold expression can be expressed as 110 - min(2*distance standard deviation, 30), and the corresponding speed standard deviation threshold can be 60.

[0050] Figure 4 A block diagram of an obstacle recognition system for adaptive cruise control of a vehicle according to some embodiments of the present disclosure is shown. The various embodiments in this specification are described in a progressive manner, with reference to each other for similar or identical parts. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple, and relevant parts can be referred to the description of the method embodiments. Figure 4As shown, the obstacle recognition system 400 for adaptive cruise control includes an image acquisition module 402 configured to acquire road images of the vehicle when it is in adaptive cruise control mode. The obstacle recognition system 400 also includes a light determination module 404 configured to determine whether the lighting in the vehicle's environment is abnormal based on the road images. The obstacle recognition system 400 further includes an object recognition module 406 configured to identify, in response to determining abnormal lighting in the vehicle's environment, at least one confidence level corresponding to the type of an object to be identified from the road images. Furthermore, the obstacle recognition system 400 also includes an obstacle determination module 408 configured to determine whether the object to be identified is an obstacle to be processed based on the confidence level corresponding to the type of the object and at least two state parameters.

[0051] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this specification are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0052] Figure 5 Block diagrams of electronic devices that can implement various embodiments of the present disclosure are shown. For example... Figure 5As shown, the electronic device 500 includes a processor 501, which can perform various appropriate actions and processes based on computer program instructions loaded into random access memory (RAM) 503 according to computer program instructions stored in read-only memory (ROM) 502. The RAM 503 may also store various programs and data required for the operation of the electronic device 500. The processor 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0053] The various processes and procedures described above, such as method 200, can be executed by processor 501. For example, in some embodiments, method 200 may be implemented as a software program tangibly contained in a machine-readable medium. In some embodiments, part or all of the software program may be loaded into and / or installed onto electronic device 500 via ROM 502. When the software program is loaded into RAM 503 and executed by processor 501, one or more actions of method 200 described above may be performed.

[0054] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0055] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0056] This disclosure can be a method, apparatus, system, and / or program product. The program product may include a machine-readable storage medium on which machine-readable program instructions for performing various aspects of this disclosure are loaded. The machine-readable program instructions described herein can be downloaded from the machine-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the machine-readable program instructions from the network and forwards them to the machine-readable storage medium in the respective computing / processing device.

[0057] Machine program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. Machine-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as a standalone software package, partially on a user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the machine-readable program instructions to implement various aspects of this disclosure.

[0058] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0059] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for obstacle recognition for adaptive cruise control of a vehicle, characterized in that, The method comprises: obtaining a road image of the vehicle based on the vehicle being in an adaptive cruise state; determining whether the light of the environment in which the vehicle is located is abnormal based on the road image; in response to determining that the light of the environment in which the vehicle is located is abnormal, identifying a confidence degree corresponding to a type of at least one to-be-identified object from the road image; and based on the confidence degree corresponding to the type of the to-be-identified object and at least two state parameters, determining whether the to-be-identified object is a to-be-handled obstacle. The method of determining whether the light of the environment in which the vehicle is located is abnormal based on the road image comprises:

2. The method of claim 1, wherein, performing color space conversion processing on the road image, and counting a proportion of the number of pixels corresponding to a low-light intensity interval from the road image after the color space conversion processing; determining whether the proportion of the number of pixels corresponding to the low-light intensity interval is less than a preset proportion threshold value; in response to determining that the proportion of the number of pixels corresponding to the low-light intensity interval is less than the preset proportion threshold value, determining that the light of the environment in which the vehicle is located is normal; or in response to determining that the proportion of the number of pixels corresponding to the low-light intensity interval is not less than the preset proportion threshold value, determining that the light of the environment in which the vehicle is located is abnormal. The method of determining whether the to-be-identified object is a to-be-handled obstacle based on the confidence degree corresponding to the type of the to-be-identified object and at least two state parameters comprises:

3. The method of claim 2, wherein, determining whether the confidence degree corresponding to the type of the to-be-identified object is less than a preset confidence degree threshold value; in response to determining that the confidence degree corresponding to the type of the to-be-identified object is not less than the preset confidence degree threshold value, determining whether the to-be-identified object is a to-be-handled obstacle based on at least two state parameters; or in response to determining that the confidence degree corresponding to the type of the to-be-identified object is less than the preset confidence degree threshold value, determining that the to-be-identified object is not a to-be-handled obstacle. The at least two state parameters comprise a current speed of the to-be-identified object, a current speed of the vehicle, a relative speed between the vehicle and the to-be-identified object, a longitudinal distance, and a distance standard deviation.

4. The method of claim 3, wherein, The method of determining whether the to-be-identified object is a to-be-handled obstacle based on at least two state parameters comprises: determining whether the longitudinal distance is less than a distance threshold value corresponding to the type of the to-be-identified object; in response to determining that the longitudinal distance is not less than the distance threshold value corresponding to the type of the to-be-identified object, determining that the to-be-identified object is a to-be-handled obstacle; or in response to determining that the longitudinal distance is less than the distance threshold value corresponding to the type of the to-be-identified object, determining whether the current speed of the to-be-identified object is less than a preset vehicle speed threshold value; in response to determining that the current speed of the to-be-identified object is less than the preset vehicle speed threshold value, determining whether the to-be-identified object is a to-be-handled obstacle based on the current speed of the vehicle, the type of the to-be-identified object, the current speed of the to-be-identified object, the relative speed between the vehicle and the to-be-identified object, the longitudinal distance, and the distance standard deviation. ​ 5. The method of claim 4, wherein, The determining whether the to-be-identified object is the to-be-handled obstacle based on the at least two state parameters further includes: In response to determining that the current speed of the to-be-identified object is not less than the preset vehicle speed threshold, determining that the to-be-identified object is the to-be-handled obstacle.

6. The method of claim 4, wherein, The determining whether the to-be-identified object is the to-be-handled obstacle based on the current speed of the vehicle, the type of the to-be-identified object, the current speed of the to-be-identified object, the relative speed between the vehicle and the to-be-identified object, the longitudinal distance, and the distance standard deviation includes: determining a distance standard deviation threshold expression of the to-be-identified object based on the current speed of the vehicle and the type of the to-be-identified object, and determining a target distance standard deviation threshold based on the distance standard deviation and the distance standard deviation threshold expression; determining an actual distance standard deviation based on the longitudinal distance, and determining whether the actual distance standard deviation is less than the target distance standard deviation threshold; in response to determining that the actual distance standard deviation is not less than the target distance standard deviation threshold, determining a target speed standard deviation threshold of the to-be-identified object based on the current speed of the vehicle and the type of the to-be-identified object; determining an actual speed standard deviation of the to-be-identified object based on the current speed of the vehicle, the current speed of the to-be-identified object, the relative speed, and the longitudinal distance, and determining whether the actual speed standard deviation is less than the target speed standard deviation threshold; and in response to determining that the actual speed standard deviation is not less than the target speed standard deviation threshold, determining that the to-be-identified object is the to-be-handled obstacle.

7. The method of claim 6, wherein, The determining whether the to-be-identified object is the to-be-handled obstacle based on the current speed of the vehicle, the type of the to-be-identified object, the current speed of the to-be-identified object, the relative speed between the vehicle and the to-be-identified object, the longitudinal distance, and the distance standard deviation further includes: in response to determining that the actual distance standard deviation is less than the target distance standard deviation threshold, determining that the to-be-identified object is not the to-be-handled obstacle; or in response to determining that the actual speed standard deviation is less than the target speed standard deviation threshold, determining that the to-be-identified object is not the to-be-handled obstacle.

8. The method according to any one of claims 1 to 7, characterized in that, The method further includes: in response to determining that the to-be-identified object is the to-be-handled obstacle, performing filtering processing on the to-be-handled obstacle.

9. A barrier recognition system for adaptive cruise control of a vehicle, characterized in that comprising: an image acquisition module configured to acquire a road image of a vehicle based on the vehicle being in an adaptive cruise state; a light judgment module configured to determine whether the light of an environment in which the vehicle is located is abnormal based on the road image; an object identification module configured to identify a confidence level corresponding to a type of at least one to-be-identified object from the road image in response to determining that the light of the environment in which the vehicle is located is abnormal; and an obstacle determination module configured to determine whether the to-be-identified object is a to-be-handled obstacle based on the confidence level corresponding to the type of the to-be-identified object and at least two state parameters. comprising:

10. An electronic device, comprising: one or more processors, and ​ a memory associated with the one or more processors, the memory for storing program instructions that, when read and executed by the one or more processors, perform the steps of the method of any of claims 1-8.

Images