Obstacle filtering method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of automatic driving, and particularly discloses an obstacle filtering method and device, computer equipment and a storage medium. First position information, corresponding to a first detection frame, of a target vehicle and second position information, corresponding to a second detection frame, of a candidate obstacle are obtained; then the overlapping degree of the first detection frame and the second detection frame is determined through the first position information and the second position information; and finally, filtering the candidate obstacles with the overlapping degree reaching the preset overlapping level as to-be-filtered obstacles. The error degree of candidate obstacle recognition is judged by determining the overlapping degree of the first detection frame and the second detection frame, and therefore whether wrong recognition of the current candidate obstacle will cause wrong issuing of a driving instruction or not is determined; and if the overlapping degree reaches the preset overlapping level, determining the candidate obstacle as a to-be-filtered obstacle and filtering the to-be-filtered obstacle, so that wrong issuing of the driving instruction is avoided, and the stability and safety of the automatic driving process are ensured.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to an obstacle filtering method, device, computer equipment, and storage medium. Background Art

[0002] With the continuous development of internet technology, autonomous driving is gradually becoming part of people's lives. Autonomous driving involves analyzing and calculating identified obstacles during a vehicle's driving process to determine a route, and then providing corresponding driving instructions based on that route.

[0003] Obstacle recognition, which includes identifying the size and location of obstacles, is a crucial step in the autonomous driving process. Faulty obstacle recognition can lead to incorrect driving instructions, causing traffic accidents and reducing driving safety. In particular, when large obstacles near the vehicle are misidentified, the obstacle's detection frame often overlaps with the autonomous vehicle's. This can cause the autonomous vehicle to brake incorrectly, reducing driving safety and stability. Summary of the Invention

[0004] In order to solve the above technical problems, the present disclosure provides an obstacle filtering method, apparatus, computer equipment and storage medium.

[0005] In a first aspect, the present invention provides an obstacle filtering method, the method comprising:

[0006] Obtain first position information of the target vehicle corresponding to the first detection frame and second position information of the candidate obstacle corresponding to the second detection frame; determine the degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information; and determine the candidate obstacle as an obstacle to be filtered when the degree of overlap reaches a preset overlap level.

[0007] The obstacle filtering method provided in this embodiment first obtains first position information of a target vehicle corresponding to a first detection frame and second position information of a candidate obstacle corresponding to a second detection frame. The method then determines the degree of overlap between the first and second detection frames based on the first and second position information. Finally, candidate obstacles whose overlap reaches a preset level are filtered as obstacles to be filtered. This embodiment determines the degree of error in candidate obstacle identification by determining the degree of overlap between the first and second detection frames, and uses this to determine whether incorrect identification of the current candidate obstacle will result in the incorrect issuance of a driving instruction. If the degree of overlap reaches the preset level, the candidate obstacle is determined as an obstacle to be filtered and filtered, thereby preventing the incorrect issuance of driving instructions and ensuring the stability and safety of the autonomous driving process.

[0008] In an optional implementation, obtaining second position information of the candidate obstacle corresponding to the second detection frame includes:

[0009] Obtaining description information corresponding to the initial obstacle, the description information including position information of a detection frame to which the initial obstacle belongs and obstacle attributes corresponding to the initial obstacle; determining whether the initial obstacle meets preset conditions based on the description information, wherein the preset conditions are that the initial obstacle is within a preset range centered on the target vehicle and that the obstacle attributes of the initial obstacle are preset attributes; if the initial obstacle meets the preset conditions, determining the initial obstacle as a candidate obstacle, and obtaining second position information of a second detection frame corresponding to the candidate obstacle from the description information of the initial obstacle.

[0010] In an optional implementation, obtaining description information corresponding to the initial obstacle includes:

[0011] Obtain image data at the current moment; input the image data into a pre-built bird's-eye view prediction model and output description information corresponding to the obstacles to be processed; based on the description information corresponding to the obstacles to be processed, use a non-maximum suppression algorithm to post-process the obstacles to be processed to determine the initial obstacles from the obstacles to be processed and obtain the description information corresponding to the initial obstacles.

[0012] The obstacle filtering method provided in this embodiment uses the NMS algorithm to post-process pending obstacles, removes redundant pending obstacles, and selects initial obstacles with optimal bounding boxes. This reduces the amount of data to a certain extent, allowing for rapid selection of candidate obstacles from the initial obstacles, thereby improving work efficiency.

[0013] In an optional implementation, determining the degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information includes:

[0014] Based on the second position information, determine the center point coordinates of the second detection frame; based on the center point coordinates and the first position information, determine whether the center point coordinates are within the area covered by the first detection frame; and determine the degree of overlap between the first detection frame and the second detection frame based on the judgment result.

[0015] In an optional implementation, determining the degree of overlap between the first detection frame and the second detection frame according to the judgment result includes:

[0016] When the center point coordinates are within the area covered by the first detection frame, the degree of overlap is determined to be high overlap; when the center point coordinates are not within the area covered by the first detection frame, based on the first position information and the second position information, the proportion of the overlapping area of ​​the first detection frame and the second detection frame to the corresponding area of ​​the first detection frame is calculated; and the degree of overlap is determined based on the proportion.

[0017] In an optional implementation, determining the degree of overlap based on the proportion includes:

[0018] The proportion is compared with a preset proportion threshold; when the proportion is greater than or equal to the preset proportion threshold, the degree of overlap is determined to be moderate overlap; when the proportion is less than the preset proportion threshold, the degree of overlap is determined to be slight overlap.

[0019] The obstacle filtering method provided in this embodiment first determines the high overlap situation through the coordinates of the center point of the second detection frame, and then determines the medium overlap and low overlap by the proportion of the overlapping area in the first detection frame. The above method can more accurately divide the degree of overlap, so that the final filtering result is more accurate.

[0020] In an optional embodiment, the method further includes:

[0021] When the degree of overlap is slight, the candidate obstacle is determined as the target obstacle.

[0022] In a second aspect, the present invention provides an obstacle filtering device, comprising:

[0023] An acquisition module is configured to obtain first position information of the target vehicle corresponding to the first detection frame and second position information of the candidate obstacle corresponding to the second detection frame. A determination module is configured to determine the degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information. A filtering module is configured to determine the candidate obstacle as an obstacle to be filtered when the degree of overlap reaches a preset overlap level.

[0024] In a third aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the obstacle filtering method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0025] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the obstacle filtering method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0027] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0028] Figure 1 A flowchart of an obstacle filtering method provided in an embodiment of the present disclosure;

[0029] Figure 2 A flowchart of another obstacle filtering method provided by an embodiment of the present disclosure;

[0030] Figure 3 A flowchart of another obstacle filtering method provided in an embodiment of the present disclosure;

[0031] Figure 4 A structural block diagram of an obstacle filtering device provided in an embodiment of the present disclosure;

[0032] Figure 5 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0034] The following description sets forth many specific details to facilitate a full understanding of the present disclosure. However, the present disclosure may also be implemented in other ways than those described herein. It is apparent that the embodiments described in the specification are only some of the embodiments of the present disclosure, not all of them. All other embodiments derived by those skilled in the art based on the embodiments of the present disclosure without inventive effort are also within the scope of protection of the present disclosure.

[0035] In the field of autonomous driving, safety is a primary concern for users and a key area in which current autonomous driving technology requires further advancement. Safety during autonomous driving is closely related to the accuracy of obstacle recognition. Obstacle recognition here primarily refers to the identification of obstacle location and spatial size. Errors in obstacle recognition can lead to incorrect route planning, causing the control system to issue driving instructions that are inconsistent with the current situation, potentially causing traffic accidents. Although obstacle recognition technology has reached a relatively mature stage, errors in recognition are still unavoidable, and sometimes the identified obstacle differs significantly from the actual situation. Especially when a large obstacle is near a vehicle, the speed at which the obstacle is moving often leads to significant overlap between the identified obstacle detection frame and the autonomous vehicle's detection frame. This can cause the autonomous vehicle to brake incorrectly, reducing driving safety and stability. To address the above issues, the present invention provides an obstacle filtering method, apparatus, computer device, and storage medium.

[0036] According to an embodiment of the present invention, an embodiment of an obstacle filtering method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0037] In this embodiment, an obstacle filtering method is provided. Figure 1 FIG. 1 is a flow chart of an obstacle filtering method according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0038] Step S101 : obtaining first position information of a target vehicle corresponding to a first detection frame and second position information of a candidate obstacle corresponding to a second detection frame.

[0039] Specifically, the first position information and the second position information can be obtained from the bird's-eye-view (BEV) corresponding to the current moment. The bird's-eye-view is a three-dimensional image based on the vehicle body coordinate system generated based on the information of all obstacles around the vehicle (the vehicle body coordinate system is a three-dimensional coordinate system with the autonomous driving vehicle as the coordinate origin). The bird's-eye-view includes the detection frames corresponding to the autonomous driving vehicle and all obstacles identified at the current moment, as well as the attribute information and position information of each detection frame. The attribute information is the attribute of the obstacle, such as the attribute information can be a truck, a van, a tree, a person, etc. The position information is the relevant information used to identify the position of the detection frame, such as the coordinates of each vertex of the detection frame, the coordinates of the center point of the detection frame, and the length information of each side of the detection frame.

[0040] Specifically, upon receiving an acquisition instruction, the first position information of the target vehicle corresponding to the first detection frame and the second position information of the candidate obstacle corresponding to the second detection frame in the BEV image can be acquired. It should be noted that the target vehicle is the autonomous vehicle, and the candidate obstacle is an obstacle in the BEV image that meets a preset condition. The preset condition is the position information of an obstacle that is likely to cause the autonomous vehicle to issue an erroneous driving instruction. For example, it can be an obstacle with attribute M within x meters in front of the target vehicle. The candidate obstacle can be one or more. The preset conditions for screening candidate obstacles are not specifically limited here, and those skilled in the art can set them according to actual needs.

[0041] Step S102 : determining the degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information.

[0042] Specifically, the degree of overlap indicates the degree of error in obstacle identification. A higher degree of overlap indicates a greater discrepancy between the identified obstacle and the actual situation, making it more likely that an erroneous driving instruction will be issued. Conversely, a lower degree of overlap indicates a smaller discrepancy between the identified obstacle and the actual situation, reducing the probability of an erroneous driving instruction being issued. There are various ways to determine the degree of overlap based on the first and second position information, which are not specifically limited here. Persons skilled in the art may choose any feasible method.

[0043] For example, in an optional embodiment, the relative positional relationship between the first detection frame and the second detection frame can be determined based on the first position information and the second position information, and the degree of overlap between the first detection frame and the second detection frame can be determined based on the relative positional relationship. The relative positional relationship can be measured by whether the center point of the second detection frame is within the first detection frame.

[0044] Exemplarily, in another optional embodiment, the overlapping area between the first detection frame and the second detection frame can be determined based on the first position information and the second position information; the degree of overlap is determined based on the proportion of the overlapping area in the area where the first detection frame is located.

[0045] Step S103: When the degree of overlap reaches a preset overlap level, the candidate obstacle is determined as an obstacle to be filtered.

[0046] Specifically, the preset overlap level is the lowest level that is likely to result in incorrect driving instructions being issued. In this embodiment, the overlap levels can include slight overlap, moderate overlap, and high overlap, and the preset overlap level can be moderate overlap. That is, when the overlap level is moderate or high, it indicates that the current candidate obstacle differs significantly from the actual situation, which could result in incorrect driving instructions being issued. In this case, the candidate obstacle is determined as a to-be-filtered obstacle and filtered out. That is, the second detection frame and related information corresponding to the to-be-filtered obstacle are deleted from the bird's-eye view. When the overlap level is slight, it indicates that the current candidate obstacle differs slightly from the actual situation, and the error is within an acceptable range. In this case, the candidate obstacle is determined as a target obstacle and retained.

[0047] The obstacle filtering method provided in this embodiment first obtains first position information of a target vehicle corresponding to a first detection frame and second position information of a candidate obstacle corresponding to a second detection frame. The method then determines the degree of overlap between the first and second detection frames based on the first and second position information. Finally, candidate obstacles whose overlap reaches a preset level are filtered as obstacles to be filtered. This embodiment determines the degree of error in candidate obstacle identification by determining the degree of overlap between the first and second detection frames, and uses this to determine whether incorrect identification of the current candidate obstacle will result in the incorrect issuance of a driving instruction. If the degree of overlap reaches the preset level, the candidate obstacle is determined as an obstacle to be filtered and filtered, thereby preventing the incorrect issuance of driving instructions and ensuring the stability and safety of the autonomous driving process.

[0048] In an alternative embodiment, if Figure 2 As shown, obtaining the second position information of the candidate obstacle corresponding to the second detection frame in step S101 includes the following steps:

[0049] Step S1011: Obtain description information corresponding to the initial obstacle.

[0050] Specifically, the description information includes the location information of the initial obstacle corresponding to the detection frame and the obstacle attributes corresponding to the initial obstacle. For detailed descriptions of the location information and obstacle attributes, please refer to Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0051] In some optional implementations, the above step S1011 includes:

[0052] Step a1: Acquire image data at the current moment.

[0053] Specifically, the image data at the current moment is the image collected by various cameras installed around the autonomous driving vehicle at the current moment.

[0054] Step a2: input the image data into the pre-built bird's-eye view prediction model and output the description information corresponding to the obstacle to be processed.

[0055] Specifically, the image data is input into the bird's-eye view prediction model, which extracts and analyzes the image data to generate a bird's-eye view. The model then outputs the descriptions of all obstacles in the bird's-eye view (i.e., obstacles to be processed) for post-processing based on the output descriptions. It should be noted that since the image data is collected while the vehicle is in motion, especially at high speeds, it is easy for one obstacle to be identified as multiple during the obstacle recognition process. Therefore, the number of obstacles in the bird's-eye view (i.e., obstacles to be processed) is often greater than the actual number of obstacles.

[0056] In step a3, based on the description information corresponding to the obstacles to be processed, the obstacles to be processed are post-processed using a non-maximum suppression algorithm to determine an initial obstacle from the obstacles to be processed and obtain the description information corresponding to the initial obstacle.

[0057] Specifically, the Non-Maximum Suppression (NMS) algorithm is a technology used for target detection that aims to select the best bounding box from a set of overlapping boxes. In this embodiment, the obstacles to be processed are post-processed using the NMS algorithm based on the description information of the detection box corresponding to the obstacle to be processed, so as to filter out the initial obstacles from the obstacles to be processed. For example, obstacle A in the actual environment is identified as three obstacles to be processed A1, A2, and A3 in the bird's-eye view. After post-processing using the NMS algorithm, A3 is filtered out as the obstacle corresponding to the best bounding box, and A3 is determined as the initial obstacle. It should be emphasized that after post-processing, the number of initial obstacles obtained is less than the number of obstacles to be processed, and is greater than or equal to the actual number of obstacles.

[0058] The obstacle filtering method provided in this embodiment uses the NMS algorithm to post-process pending obstacles, removes redundant pending obstacles, and selects initial obstacles with optimal bounding boxes. This reduces the amount of data to a certain extent, allowing for rapid selection of candidate obstacles from the initial obstacles, thereby improving work efficiency.

[0059] Step S1012: Based on the description information, determine whether the initial obstacle meets the preset conditions.

[0060] Specifically, the preset condition is that the initial obstacle is within a preset range centered on the target vehicle and the obstacle attribute of the initial obstacle is a preset attribute. For example, the preset condition is that the obstacle is within a range of 1 meter from the target vehicle and the obstacle is a truck.

[0061] Step S1013 : When the initial obstacle meets a preset condition, the initial obstacle is determined as a candidate obstacle, and second position information of the candidate obstacle corresponding to the second detection frame is obtained from the description information of the initial obstacle.

[0062] In this embodiment, an obstacle filtering method is provided. Figure 3 FIG. 1 is a flow chart of an obstacle filtering method according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0063] Step S301: Obtain the first position information of the target vehicle corresponding to the first detection frame and the second position information of the candidate obstacle corresponding to the second detection frame. Figure 1 Step S101 of the embodiment shown, and Figure 2 All the steps of the illustrated embodiment will not be repeated here.

[0064] Step S302 : determining the degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information.

[0065] Specifically, the above step S302 includes:

[0066] Step S3021: Determine the center point coordinates of the second detection frame according to the second position information.

[0067] Specifically, when the second position information is the coordinates of each vertex of the second detection frame, the coordinates of the center point of the second detection frame are determined based on the coordinates of each vertex.

[0068] Step S3022: Based on the center point coordinates and the first position information, determine whether the center point coordinates are within the area covered by the first detection frame.

[0069] Specifically, first, the area covered by the first detection frame is determined according to the first position information, and then it is determined whether the coordinates of the center point are within the area covered by the first detection frame.

[0070] Step S3023: Determine the degree of overlap between the first detection frame and the second detection frame according to the judgment result.

[0071] In some optional implementations, the above step S3023 includes:

[0072] In step b1, when the center point coordinates are within the area covered by the first detection frame, the degree of overlap is determined to be high overlap.

[0073] Step b2: When the center point coordinates are not within the area covered by the first detection frame, based on the first position information and the second position information, calculate the proportion of the overlapping area of ​​the first detection frame and the second detection frame in the corresponding area of ​​the first detection frame.

[0074] Specifically, when the coordinates of the center point are not within the coverage range of the first detection frame, the coverage area of ​​the first detection frame is first determined based on the first position information, and the coverage area of ​​the second detection frame is determined based on the second position information; then, the overlapping area of ​​the first detection frame and the second detection frame is determined based on the coverage area of ​​the first detection frame and the coverage area of ​​the second detection frame; finally, the proportion of the overlapping area to the coverage area of ​​the first detection frame (i.e., the proportion) is calculated.

[0075] Step b3: Determine the degree of overlap based on the proportion.

[0076] Specifically, the ratio is compared with a preset ratio threshold. If the ratio is greater than or equal to the preset ratio threshold, the overlap is determined to be moderate; if the ratio is less than the preset ratio threshold, the overlap is determined to be slight. The preset ratio threshold is the maximum acceptable ratio that does not cause erroneous driving instructions while ensuring safe driving.

[0077] Step S303: When the degree of overlap reaches a preset overlap level, the candidate obstacle is determined as an obstacle to be filtered. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0078] The obstacle filtering method provided in this embodiment first determines the high overlap situation through the coordinates of the center point of the second detection frame, and then determines the medium overlap and low overlap by the proportion of the overlapping area in the first detection frame. The above method can more accurately divide the degree of overlap, so that the final filtering result is more accurate.

[0079] This embodiment also provides an obstacle filtering device for implementing the above-mentioned embodiments and preferred embodiments. Details already described will not be repeated. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0080] This embodiment provides an obstacle filtering device, such as Figure 4 As shown, including:

[0081] The acquisition module 401 is used to acquire first position information of the target vehicle corresponding to the first detection frame and second position information of the candidate obstacle corresponding to the second detection frame.

[0082] The determination module 402 is configured to determine the degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information.

[0083] The filtering module 403 is configured to determine the candidate obstacle as an obstacle to be filtered when the degree of overlap reaches a preset overlap level.

[0084] In some optional implementations, the acquisition module 401 includes:

[0085] The acquisition submodule is configured to obtain descriptive information corresponding to the initial obstacle, including positional information of a detection frame to which the initial obstacle belongs and obstacle attributes corresponding to the initial obstacle. The first judgment submodule is configured to determine, based on the descriptive information, whether the initial obstacle meets preset conditions, where the preset conditions are that the initial obstacle is within a preset range centered on the target vehicle and that the obstacle attributes of the initial obstacle are preset attributes. The first determination submodule is configured to determine the initial obstacle as a candidate obstacle when the preset conditions are met, and to obtain second positional information of a second detection frame corresponding to the candidate obstacle from the descriptive information of the initial obstacle.

[0086] In some optional implementations, the acquisition submodule includes:

[0087] The acquisition unit is used to acquire image data at the current moment; input the image data into a pre-built bird's-eye view prediction model and output descriptive information corresponding to the obstacle to be processed; and the post-processing unit is used to post-process the obstacle to be processed using a non-maximum suppression algorithm based on the descriptive information corresponding to the obstacle to be processed, so as to determine the initial obstacle from the obstacles to be processed and obtain the descriptive information corresponding to the initial obstacle.

[0088] In some optional implementations, the determining module 402 includes:

[0089] The second determination submodule is used to determine the center point coordinates of the second detection frame based on the second position information; the second judgment submodule is used to judge whether the center point coordinates are within the area covered by the first detection frame based on the center point coordinates and the first position information; the third determination submodule is used to determine the degree of overlap between the first detection frame and the second detection frame based on the judgment result.

[0090] In some optional implementations, the third determining submodule includes:

[0091] The first determination unit is used to determine the degree of overlap as high overlap when the center point coordinates are within the area covered by the first detection frame; the calculation unit is used to calculate the proportion of the overlapping area of ​​the first detection frame and the second detection frame to the corresponding area of ​​the first detection frame based on the first position information and the second position information when the center point coordinates are not within the area covered by the first detection frame; the second determination unit is used to determine the degree of overlap based on the proportion.

[0092] In some optional implementations, the second determining unit includes:

[0093] A comparison subunit is used to compare the proportion with a preset proportion threshold; a first determination subunit is used to determine the degree of overlap as moderate overlap when the proportion is greater than or equal to the preset proportion threshold; and a second determination subunit is used to determine the degree of overlap as slight overlap when the proportion is less than the preset proportion threshold.

[0094] In some optional embodiments, the device further comprises:

[0095] The retention module is used to determine the candidate obstacle as the target obstacle when the overlap degree is slight.

[0096] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0097] The obstacle filtering device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0098] The embodiment of the present invention also provides a computer device having the above Figure 4 Obstruction filtering device shown.

[0099] See also Figure 5 , Figure 5 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 5 As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 10 is taken as an example.

[0100] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0101] The memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0102] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0103] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0104] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0105] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0106] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0107] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. An obstacle filtering method, characterized in that: The method comprises: Obtain first position information of the target vehicle corresponding to the first detection frame and second position information of the candidate obstacle corresponding to the second detection frame; determining, based on the first position information and the second position information, a degree of overlap between the first detection frame and the second detection frame; When the degree of overlap reaches a preset overlap level, the candidate obstacle is determined as an obstacle to be filtered.

2. The obstacle filtering method according to claim 1, characterized in that: Obtaining the second position information of the candidate obstacle corresponding to the second detection frame includes: Obtaining description information corresponding to the initial obstacle, the description information including position information of the detection frame to which the initial obstacle belongs and obstacle attributes corresponding to the initial obstacle; Based on the description information, determining whether the initial obstacle meets a preset condition, wherein the preset condition is that the initial obstacle is within a preset range centered on the target vehicle and the obstacle attribute of the initial obstacle is a preset attribute; When the initial obstacle meets the preset condition, the initial obstacle is determined as the candidate obstacle, and second position information of the candidate obstacle corresponding to the second detection frame is obtained from the description information of the initial obstacle.

3. The obstacle filtering method according to claim 2, characterized in that: The obtaining of description information corresponding to the initial obstacle includes: Get the image data at the current moment; Input the image data into a pre-built bird's-eye view prediction model and output description information corresponding to the obstacle to be processed; Based on the description information corresponding to the obstacles to be processed, the obstacles to be processed are post-processed using a non-maximum suppression algorithm to determine the initial obstacle from the obstacles to be processed and obtain the description information corresponding to the initial obstacle.

4. The obstacle filtering method according to any one of claims 1 to 3, characterized in that: The determining, based on the first position information and the second position information, a degree of overlap between the first detection frame and the second detection frame includes: Determining the center coordinates of the second detection frame according to the second position information; Based on the center point coordinates and the first position information, determining whether the center point coordinates are within an area covered by the first detection frame; The degree of overlap between the first detection frame and the second detection frame is determined according to the judgment result.

5. The obstacle filtering method according to claim 4, characterized in that: The determining, according to the judgment result, the degree of overlap between the first detection frame and the second detection frame includes: When the center point coordinates are within the area covered by the first detection frame, determining the degree of overlap as high overlap; When the center point coordinates are not within the area covered by the first detection frame, calculating, based on the first position information and the second position information, a proportion of the overlapping area of ​​the first detection frame and the second detection frame in the area corresponding to the first detection frame; The degree of overlap is determined based on the proportion.

6. The obstacle filtering method according to claim 5, characterized in that: The determining the degree of overlap based on the proportion includes: Comparing the proportion with a preset proportion threshold; When the proportion is greater than or equal to the preset proportion threshold, the degree of overlap is determined to be moderate overlap; When the proportion is less than the preset proportion threshold, the overlap degree is determined to be slight overlap.

7. The obstacle filtering method according to claim 1, characterized in that: The method further comprises: When the degree of overlap is slight, the candidate obstacle is determined as a target obstacle.

8. An obstacle filtering device, characterized in that: The device comprises: An acquisition module, configured to acquire first position information of the target vehicle corresponding to the first detection frame and second position information of the candidate obstacle corresponding to the second detection frame; a determination module, configured to determine a degree of overlap between the first detection frame and the second detection frame based on the first position information and the second position information; The filtering module is configured to determine the candidate obstacle as an obstacle to be filtered when the degree of overlap reaches a preset overlap level.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the obstacle filtering method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the obstacle filtering method according to any one of claims 1 to 7.

Images