CN112859810A - ADAS algorithm verification method and device based on Carla platform - Google Patents
ADAS algorithm verification method and device based on Carla platform Download PDFInfo
- Publication number
- CN112859810A CN112859810A CN202110041907.2A CN202110041907A CN112859810A CN 112859810 A CN112859810 A CN 112859810A CN 202110041907 A CN202110041907 A CN 202110041907A CN 112859810 A CN112859810 A CN 112859810A
- Authority
- CN
- China
- Prior art keywords
- tested
- platform
- adas algorithm
- test
- adas
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0208—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the configuration of the monitoring system
- G05B23/0213—Modular or universal configuration of the monitoring system, e.g. monitoring system having modules that may be combined to build monitoring program; monitoring system that can be applied to legacy systems; adaptable monitoring system; using different communication protocols
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an ADAS algorithm verification method and device based on Carla platform. The method comprises the following steps: acquiring information of an object to be tested and information of an environment to be tested, and creating a corresponding simulation object in a Carla platform according to the information of the object to be tested and the information of the environment to be tested; connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame; processing the test picture image frame through an ADAS algorithm, and feeding back a processing result to a Carla platform real-time picture; and obtaining an expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result. The invention can meet different environmental requirements of the ADAS algorithm by means of the Carla platform, creates a specific test environment, tests the performance of the algorithm in multiple aspects, can improve the test performance of the ADAS algorithm and improves the robustness of the ADAS algorithm.
Description
Technical Field
The invention relates to the technical field of computer software, in particular to an ADAS algorithm verification method and device based on Carla platform.
Background
In the era of artificial intelligence and fire and heat, the unmanned driving algorithm based on deep learning is infinite, the model training of the deep learning algorithm and the algorithm robustness effect verification can not be independent of the establishment and the test of a simulation environment, the environment of the field of unmanned driving in reality is usually complex, and the specific environment of the algorithm to be tested is difficult to simulate, so that the establishment of the specific test environment for the algorithm for testing the effect of the algorithm is an indispensable step for improving the performance of the deep learning algorithm.
The ADAS algorithm mainly relates to two major aspects of target collision early warning and lane line deviation early warning, road information pictures or videos obtained in real time are adopted in the current ADAS algorithm test, environments built in practice like vehicle collision early warning and lane line deviation early warning are simple, the environments encountered in practical application are far more complex than test environments, higher requirements are undoubtedly put forward on the performance of the algorithm, and how to build the test environments aiming at different algorithms becomes a difficult problem for ADAS research field personnel. The traditional test environment is simple, the cost is high, the test effect is single, and much uncertainty exists. Therefore, an ADAS algorithm verification method based on carra platform is needed to improve the algorithm performance.
The above is only for the purpose of assisting understanding of the technical aspects of the present invention, and does not represent an admission that the above is prior art.
Disclosure of Invention
In view of this, the invention provides an ADAS algorithm verification method and device based on a carra platform, and aims to solve the technical problem that the ADAS algorithm test performance cannot be improved through the carra platform in the prior art.
The technical scheme of the invention is realized as follows:
in one aspect, the present invention provides an ADAS algorithm verification method based on carra platform, which comprises the following steps:
s1, acquiring information of the object to be tested and information of the environment to be tested, and creating a corresponding simulation object in the Carla platform according to the information of the object to be tested and the information of the environment to be tested;
s2, connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame;
s3, processing the image frame of the test picture through an ADAS algorithm, and feeding back the processing result to a real-time picture of a Carla platform;
s4, obtaining the expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result.
Based on the above technical solution, preferably, in step S1, obtaining information of an object to be tested and information of an environment to be tested, and creating a corresponding simulation object in the carala platform according to the information of the object to be tested and the information of the environment to be tested, further comprising the following steps of obtaining information of the object to be tested and the information of the environment to be tested, where the information of the object to be tested includes: pedestrian information and car information, the environmental information that awaits measuring includes: and creating a corresponding simulation object in the Carla platform according to the weather information and the road condition information and the information of the object to be tested and the environment information to be tested.
On the basis of the above technical solution, preferably, in step S2, the simulation object is connected to the ADAS algorithm to obtain an operation instruction to be tested, the simulation object is controlled by the operation instruction to be tested, and a test picture image frame is obtained, and the method further includes the following steps of constructing a python image interface, and the ADAS algorithm is connected to the simulation picture corresponding to the simulation object by the python image interface to obtain the operation instruction to be tested, where the operation instruction to be tested includes: and a collision early warning test flow and a lane departure test flow, wherein the simulation object is controlled by the operation instruction to be tested, and a test picture image frame is obtained.
On the basis of the technical scheme, preferably, the method comprises the steps of extracting a vehicle operation instruction, camera attributes and test target data from the operation instruction to be tested, constructing a test picture according to the operation instruction to be tested, projecting the simulation object into the test picture, and controlling the simulation object through a control strategy to obtain the test picture image frame.
On the basis of the above technical solution, preferably, in step S3, the ADAS algorithm is used to process the test image frame, and the processing result is fed back to the real-time image of the cara platform, and the method further includes the following steps, when the operation instruction to be tested is a collision warning test process, the ADAS algorithm is used to process the test image frame, obtain the relative position between the simulation object and the detection target, set the relative position warning range and the corresponding warning behavior, compare the relative position with the relative position warning range, and obtain the corresponding warning behavior according to the comparison result, where the warning behavior includes: warning, deceleration, stop, detour and collision, and feeding back the early warning behavior as a processing result to a Carla platform real-time picture.
On the basis of the above technical solution, preferably, in step S3, the ADAS algorithm is used to process the test image frame, and the processing result is fed back to the real-time image of the cara platform, and the method further includes the following steps that when the operation instruction to be tested is a lane departure test procedure, the ADAS algorithm is used to process the test image frame, obtain the relative angle between the simulation object and the detection target, set the relative angle warning range and the corresponding warning behavior, compare the relative angle with the relative angle warning range, and obtain the corresponding warning behavior according to the comparison result, where the warning behavior includes: and left tilting and right tilting, and feeding the early warning behavior as a processing result back to a real-time picture of the Carla platform.
On the basis of the above technical solution, preferably, in step S4, obtaining an expected testing effect of the ADAS algorithm, comparing the processing result with the expected testing effect, and verifying the ADAS algorithm according to the comparison result, further comprising the steps of obtaining the expected testing effect of the ADAS algorithm, comparing the processing result with the expected testing effect, and when the processing result satisfies the expected testing effect, determining that the ADAS algorithm reaches the expectation; and when the processing result does not meet the expected test effect, judging that the ADAS algorithm does not reach the expectation.
Still further preferably, the ADAS algorithm verifying device based on carra platform includes:
the acquisition module is used for acquiring information of the object to be tested and the environment information to be tested, and creating a corresponding simulation object in the Carla platform according to the information of the object to be tested and the environment information to be tested;
the test module is used for connecting the simulation object with an ADAS algorithm, acquiring an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested and acquiring a test picture image frame;
the processing module is used for processing the test picture image frame through an ADAS algorithm and feeding back a processing result to a Carla platform real-time picture;
and the verification module is used for acquiring the expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect and verifying the ADAS algorithm according to the comparison result.
In a second aspect, the ADAS algorithm verification method based on cara platform further includes an electronic device, where the electronic device includes: a memory, a processor and a cara platform based ADAS algorithm verification method program stored on the memory and executable on the processor, the cara platform based ADAS algorithm verification method program configured to implement the steps of the cara platform based ADAS algorithm verification method as described above.
In a third aspect, the ADAS algorithm verification method based on carra platform further includes a storage medium, where the storage medium is a computer storage medium, and the computer storage medium stores an ADAS algorithm verification method program based on carra platform, and when executed by a processor, the ADAS algorithm verification method program based on carra platform implements the steps of the ADAS algorithm verification method based on carra platform.
Compared with the prior art, the ADAS algorithm verification method based on Carla platform has the following beneficial effects:
(1) the establishment of the specific algorithm environment can meet different environment requirements of the ADAS algorithm, test environment objects such as test weather attributes, test vehicle attributes and test scenes can be changed by means of the Carla platform, a specific test environment is created, the performance of the algorithm can be tested in multiple aspects, and meanwhile the performance of the algorithm can be improved.
(2) Through the object established with the help of Carla platform, can more convenient and fast set up the environment, the test object in the most real environment all can find the substitute object in Carla to greatly practice thrift the cost spending and the time of setting up the environment, make the algorithm test verify more with swift, convenient.
(3) By means of the test environment created by the Carla platform, a complex dangerous scene which is difficult to realize in reality can be built, and the test simulation requirement of the ADAS on the aspect of driving safety performance is met.
(4) The robustness of the algorithm tested by the platform is greatly improved, chips are added for application and popularization of the ADAS algorithm in actual life, and the algorithm is a large simulation tool of researchers in the ADAS field.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic structural diagram of an electronic device in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a first embodiment of the ADAS algorithm verification method based on Carla platform according to the present invention;
FIG. 3 is a test environment screenshot constructed by collision warning of the ADAS algorithm verification method based on Carla platform according to the present invention;
FIG. 4 is a test environment screenshot of lane departure warning setup of the ADAS algorithm verification method based on Carla platform according to the present invention;
fig. 5 is a functional module diagram of the ADAS algorithm verification method based on carra platform according to the first embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
As shown in fig. 1, the electronic device may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a WIreless interface (e.g., a WIreless-FIdelity (WI-FI) interface). The Memory 1005 may be a Random Access Memory (RAM) Memory, or may be a Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration shown in fig. 1 does not constitute a limitation of the electronic device, and that in actual implementations the electronic device may include more or less components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, the memory 1005, which is a storage medium, may include an operating system, a network communication module, a user interface module, and an ADAS algorithm verification method program based on carra platform.
In the electronic device shown in fig. 1, the network interface 1004 is mainly used for establishing a communication connection between the electronic device and a server storing all data required in the system of the ADAS algorithm verification method based on the cara platform; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 of the electronic device for verifying the ADAS algorithm based on the carra platform of the present invention may be arranged in the electronic device for verifying the ADAS algorithm based on the carra platform, and the electronic device for verifying the ADAS algorithm based on the carra platform calls the program of the ADAS algorithm based on the carra platform stored in the memory 1005 through the processor 1001 and executes the method for verifying the ADAS algorithm based on the carra platform provided by the present invention.
Referring to fig. 2, fig. 2 is a schematic flow chart of a first embodiment of the ADAS algorithm verification method based on carra platform according to the present invention.
In this embodiment, the ADAS algorithm verification method based on carra platform includes the following steps:
s1: acquiring information of an object to be tested and information of an environment to be tested, and creating a corresponding simulation object in a Carla platform according to the information of the object to be tested and the information of the environment to be tested.
It should be understood that, in the present embodiment, the system obtains information of an object to be tested and information of an environment to be tested, where the information of the object to be tested includes: pedestrian information and car information, the environmental information that awaits measuring includes: and creating a corresponding simulation object in the Carla platform according to the weather information and the road condition information and the information of the object to be tested and the environment information to be tested.
S2: and connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame.
It should be understood that the system will then construct a python image interface, and connect the ADAS algorithm with the simulation picture corresponding to the simulation object through the python image interface to obtain the operation instruction to be tested, where the operation instruction to be tested includes: and the collision early warning test process and the lane departure test process control the simulation object through the to-be-tested operation instruction and acquire a test picture image frame, wherein the python image interface is a function interface and is used for transmitting the image and outputting a result image.
It should be understood that the specific steps are as follows: a function interface is set up to link an ADAS algorithm with a video or an image in a Carla environment, an automobile driving mode is set through keyboard keys, a manual mode represents that an automobile can be driven to any place, and an automatic mode represents that the automobile runs according to a specified route; the method comprises the steps of setting objects such as weather, pedestrians and other vehicles in the environment after a driving mode is set, enabling the objects to be added or modified through keyboard switching, enabling the automobile to run after the objects are set, and enabling a picture displayed in the running process to be a real-time picture shot by an automobile body camera.
It should be understood that the operation instructions to be tested include: the system can construct a test picture according to the operation instruction to be tested, project the simulation object into the test picture, and control the simulation object through a control strategy to obtain a test picture image frame.
It should be understood that in a Windows or Linux system, a test script is configured according to the ADAS algorithm test function requirement, the test script is usually written in a Python form, and the writing framework thereof is mainly divided into: establishing a Carla client port, loading a Carla test road map, establishing and setting attributes of an automobile object in Carla, and establishing and configuring attributes of a vehicle body camera in Carla; the invention builds a testing script based on pygame, which is a game processing library of Python and can realize the function of manual or automatic driving by operating an automobile in Carla environment by using a keyboard.
It should be understood that an Advanced Driver Assistance System (ADAS), which is an active safety technology for collecting environmental data inside and outside a vehicle at the first time and performing technical processes such as identification, detection and tracking of static and dynamic objects by using various sensors mounted on the vehicle, so that a Driver can detect a possible danger at the fastest time to draw attention and improve safety. The ADAS uses sensors, such as cameras, radars, lasers, and ultrasonic waves, which detect light, heat, pressure, or other variables used to monitor the state of the vehicle, and are usually located in the front and rear bumpers, side-view mirrors, and the inside of the steering column or on the windshield of the vehicle.
S3: and processing the test picture image frame through an ADAS algorithm, and feeding back a processing result to a Carla platform real-time picture.
It should be understood that, when the operation instruction to be tested is a collision early warning test flow, the ADAS algorithm is used to process the test image frame, obtain the relative position between the simulation object and the detection target, set a relative position warning range and corresponding early warning behavior, compare the relative position with the relative position warning range, and obtain corresponding early warning behavior according to the comparison result, where the early warning behavior includes: warning, deceleration, stop, detour and collision, and feeding back the early warning behavior as a processing result to a Carla platform real-time picture.
It should be understood that the target collision warning process of ADAS in carra environment is as follows: carra environmental test vehicle settings, including: testing attributes such as vehicle speed, running road track and the like, and testing attributes such as vehicle size, color and style; carra environment test vehicle camera loading setting, namely setting the number of cameras, the positions of the cameras, the image resolution and other attributes; the Carla environment target detection algorithm trains model loading setting, namely the loading of the training model and the internal parameter setting; setting targets on a carra environment vehicle running route, namely setting targets such as a pedestrian target and a vehicle target; carla environment test vehicle early warning behavior is sent, namely the relative position of the test vehicle and a detection target is calculated, and warning, deceleration, stopping, detouring and other early warning behaviors are sent.
It should be understood that, as shown in fig. 3, in order to construct a test environment screenshot for collision warning, when the system detects a front automobile target, that is, a collision with an automobile is about to occur, a red trapezoid frame in the diagram is a collision area, a red rectangle frame is the detected front automobile target, and a screen displays red information: FCW: waning, which indicates that a collision is about to occur, the left small image at the upper right is a road segmentation image, and the right small image is a live view image.
It should be understood that, when the operation instruction to be tested is a lane departure test flow, the ADAS algorithm is used to process the test image frame, obtain the relative angle between the simulation object and the detection target, set a relative angle warning range and corresponding early warning behavior, compare the relative angle with the relative angle warning range, and obtain the corresponding early warning behavior according to the comparison result, where the early warning behavior includes: and left tilting and right tilting, and feeding the early warning behavior as a processing result back to a real-time picture of the Carla platform.
It should be understood that the lane departure warning procedure of ADAS in carra environment is as follows: carra environmental test vehicle settings, including: testing attributes such as vehicle speed, running road track and the like, and testing attributes such as vehicle size, color and style; carra environment test vehicle camera loading setting, namely setting the number of cameras, the positions of the cameras, the image resolution and other attributes; the Carla environment target detection algorithm trains model loading setting, namely the loading of the training model and the internal parameter setting; detecting left and right lane lines of Carla environment, namely detecting left and right lane lines of a linear road, left and right lane lines of a non-linear road, and detecting a solid line with a dotted line and a white line with a yellow line; and (5) warning the departure of the carrla environment lane line, namely judging whether the lane line meets the departure warning condition.
It should be understood that, as shown in fig. 4, in order to capture a test environment built for lane departure warning, when the vehicle turns right suddenly, the system detects that the red trapezoid frame area of the vehicle inclines right, and the real-time interface displays red information LDW: WARN _ Right, the lane departure warning is indicated, the left small image at the upper Right is a road segmentation image, and the Right small image is a live image.
S4: and obtaining an expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result.
It should be understood that, the system will obtain the expected testing effect of the ADAS algorithm, compare the processing result with the expected testing effect, and when the processing result meets the expected testing effect, judge that the ADAS algorithm reaches the expectation; and when the processing result does not meet the expected test effect, judging that the ADAS algorithm does not reach the expectation.
The above description is only for illustrative purposes and does not limit the technical solutions of the present application in any way.
As can be easily found from the above description, in the present embodiment, by acquiring information of an object to be tested and information of an environment to be tested, a corresponding simulation object is created in the cara platform according to the information of the object to be tested and the information of the environment to be tested; connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame; processing the test picture image frame through an ADAS algorithm, and feeding back a processing result to a Carla platform real-time picture; and obtaining an expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result. According to the embodiment, different environment requirements of the ADAS algorithm can be met by means of the Carla platform, a specific test environment is created, the performance of the algorithm is tested in multiple aspects, the test performance of the ADAS algorithm can be improved, and the robustness of the ADAS algorithm is improved.
In addition, the embodiment of the invention also provides an ADAS algorithm verification device based on the Carla platform. As shown in fig. 5, the ADAS algorithm verification apparatus based on cara platform includes: an acquisition module 10, a test module 20, a calculation module 30 and an indexing module 40.
The acquisition module 10 is configured to acquire information of an object to be tested and environment information to be tested, and create a corresponding simulation object in the cara platform according to the information of the object to be tested and the environment information to be tested;
the test module 20 is configured to connect the simulation object with an ADAS algorithm, obtain an operation instruction to be tested, control the simulation object through the operation instruction to be tested, and obtain a test picture image frame;
the processing module 30 is used for processing the test picture image frame through an ADAS algorithm and feeding back a processing result to a Carla platform real-time picture;
and the verification module 40 is used for acquiring an expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result.
In addition, it should be noted that the above-described embodiments of the apparatus are merely illustrative, and do not limit the scope of the present invention, and in practical applications, a person skilled in the art may select some or all of the modules to implement the purpose of the embodiments according to actual needs, and the present invention is not limited herein.
In addition, the technical details that are not described in detail in this embodiment may refer to the ADAS algorithm verification method based on carra platform provided in any embodiment of the present invention, and are not described herein again.
In addition, an embodiment of the present invention further provides a storage medium, where the storage medium is a computer storage medium, and an ADAS algorithm verification method program based on a carra platform is stored on the computer storage medium, and when executed by a processor, the ADAS algorithm verification method program based on the carra platform implements the following operations:
s1, acquiring information of the object to be tested and information of the environment to be tested, and creating a corresponding simulation object in the Carla platform according to the information of the object to be tested and the information of the environment to be tested;
s2, connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame;
s3, processing the image frame of the test picture through an ADAS algorithm, and feeding back the processing result to a real-time picture of a Carla platform;
s4, obtaining the expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result.
Further, when executed by the processor, the ADAS algorithm verification method based on carra platform further implements the following operations:
acquiring information of an object to be tested and information of an environment to be tested, wherein the information of the object to be tested comprises: pedestrian information and car information, the environmental information that awaits measuring includes: and creating a corresponding simulation object in the Carla platform according to the weather information and the road condition information and the information of the object to be tested and the environment information to be tested.
Further, when executed by the processor, the ADAS algorithm verification method based on carra platform further implements the following operations:
constructing a python image interface, connecting the ADAS algorithm with a simulation picture corresponding to a simulation object through the python image interface, and acquiring an operation instruction to be tested, wherein the operation instruction to be tested comprises: and a collision early warning test flow and a lane departure test flow, wherein the simulation object is controlled by the operation instruction to be tested, and a test picture image frame is obtained.
Further, when executed by the processor, the ADAS algorithm verification method based on carra platform further implements the following operations:
the method comprises the steps of extracting a vehicle operation instruction, camera attributes and test target data from an operation instruction to be tested, constructing a test picture according to the operation instruction to be tested, projecting a simulation object into the test picture, controlling the simulation object through a control strategy, and obtaining a test picture image frame.
Further, when executed by the processor, the ADAS algorithm verification method based on carra platform further implements the following operations:
when the operation instruction to be tested is a collision early warning test process, processing the test image frame through an ADAS algorithm, acquiring the relative position of a simulation object and a detection target, setting a relative position warning range and corresponding early warning behaviors, comparing the relative position with the relative position warning range, and acquiring the corresponding early warning behaviors according to a comparison result, wherein the early warning behaviors comprise: warning, deceleration, stop, detour and collision, and feeding back the early warning behavior as a processing result to a Carla platform real-time picture.
Further, when executed by the processor, the ADAS algorithm verification method based on carra platform further implements the following operations:
when the operation instruction to be tested is a lane departure test flow, processing the test image frame through an ADAS algorithm, acquiring a relative angle between a simulation object and a detection target, setting a relative angle warning range and corresponding early warning behaviors, comparing the relative angle with the relative angle warning range, and acquiring the corresponding early warning behaviors according to a comparison result, wherein the early warning behaviors comprise: and left tilting and right tilting, and feeding the early warning behavior as a processing result back to a real-time picture of the Carla platform.
Further, when executed by the processor, the ADAS algorithm verification method based on carra platform further implements the following operations:
obtaining an expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and judging that the ADAS algorithm achieves the expectation when the processing result meets the expected test effect; and when the processing result does not meet the expected test effect, judging that the ADAS algorithm does not reach the expectation.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. An ADAS algorithm verification method based on Carla platform is characterized in that: comprises the following steps;
s1, acquiring information of the object to be tested and information of the environment to be tested, and creating a corresponding simulation object in the Carla platform according to the information of the object to be tested and the information of the environment to be tested;
s2, connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame;
s3, processing the image frame of the test picture through an ADAS algorithm, and feeding back the processing result to a real-time picture of a Carla platform;
s4, obtaining the expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect, and verifying the ADAS algorithm according to the comparison result.
2. An ADAS algorithm verification method based on carra platform according to claim 1, characterised in that: in step S1, obtaining information of an object to be tested and environment information to be tested, and creating a corresponding simulation object in the carala platform according to the information of the object to be tested and the environment information to be tested, further comprising the following steps of obtaining information of the object to be tested and the environment information to be tested, where the information of the object to be tested includes: pedestrian information and car information, the environmental information that awaits measuring includes: and creating a corresponding simulation object in the Carla platform according to the weather information and the road condition information and the information of the object to be tested and the environment information to be tested.
3. An ADAS algorithm verification method based on carra platform according to claim 2, characterised in that: in step S2, the method includes connecting the simulation object with an ADAS algorithm to obtain an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested, and obtaining a test picture image frame, and further includes the following steps of constructing a python image interface, connecting the ADAS algorithm with a simulation picture corresponding to the simulation object through the python image interface, and obtaining the operation instruction to be tested, where the operation instruction to be tested includes: and a collision early warning test flow and a lane departure test flow, wherein the simulation object is controlled by the operation instruction to be tested, and a test picture image frame is obtained.
4. An ADAS algorithm verification method based on carra platform according to claim 3, characterised in that: the method comprises the steps of extracting a vehicle operation instruction, camera attributes and test target data from the operation instruction to be tested, constructing a test picture according to the operation instruction to be tested, projecting the simulation object into the test picture, and controlling the simulation object through a control strategy to obtain the test picture image frame.
5. An ADAS algorithm verification method based on carra platform according to claim 4, characterised in that: in step S3, the method includes processing the test picture image frame by ADAS algorithm, and feeding back the processing result to real-time picture of cara platform, and further includes the following steps, when the operation instruction to be tested is a collision warning test procedure, processing the test picture image frame by ADAS algorithm, obtaining relative position of the simulation object and the detection target, setting a relative position warning range and corresponding warning behavior, comparing the relative position with the relative position warning range, and obtaining corresponding warning behavior according to the comparison result, where the warning behavior includes: warning, deceleration, stop, detour and collision, and feeding back the early warning behavior as a processing result to a Carla platform real-time picture.
6. An ADAS algorithm verification method based on carra platform according to claim 4, characterised in that: in step S3, the method includes processing the test picture image frame by ADAS algorithm, and feeding back the processing result to real-time picture of cara platform, and further includes the following steps, when the operation instruction to be tested is a lane departure test procedure, processing the test picture image frame by ADAS algorithm, obtaining a relative angle between a simulation object and a detection target, setting a relative angle warning range and corresponding warning behavior, comparing the relative angle with the relative angle warning range, and obtaining corresponding warning behavior according to the comparison result, where the warning behavior includes: and left tilting and right tilting, and feeding the early warning behavior as a processing result back to a real-time picture of the Carla platform.
7. An ADAS algorithm verification method based on carra platform according to claim 5 or 6, characterised in that: in step S4, obtaining an expected testing effect of the ADAS algorithm, comparing the processing result with the expected testing effect, and verifying the ADAS algorithm according to the comparison result, further comprising the steps of obtaining the expected testing effect of the ADAS algorithm, comparing the processing result with the expected testing effect, and when the processing result satisfies the expected testing effect, determining that the ADAS algorithm reaches the expected result; and when the processing result does not meet the expected test effect, judging that the ADAS algorithm does not reach the expectation.
8. An ADAS algorithm verification device based on Carla platform is characterized in that the ADAS algorithm verification device based on Carla platform comprises:
the acquisition module is used for acquiring information of the object to be tested and the environment information to be tested, and creating a corresponding simulation object in the Carla platform according to the information of the object to be tested and the environment information to be tested;
the test module is used for connecting the simulation object with an ADAS algorithm, acquiring an operation instruction to be tested, controlling the simulation object through the operation instruction to be tested and acquiring a test picture image frame;
the processing module is used for processing the test picture image frame through an ADAS algorithm and feeding back a processing result to a Carla platform real-time picture;
and the verification module is used for acquiring the expected test effect of the ADAS algorithm, comparing the processing result with the expected test effect and verifying the ADAS algorithm according to the comparison result.
9. An electronic device, characterized in that the electronic device comprises: memory, a processor and a program for a cara platform based ADAS algorithm verification method stored on said memory and executable on said processor, said program for a cara platform based ADAS algorithm verification method being configured to implement the steps of the cara platform based ADAS algorithm verification method according to any of the claims 1 to 7.
10. A storage medium, characterized in that the storage medium is a computer storage medium, and the computer storage medium stores a program for an ADAS algorithm verification method based on carra platform, and the program for an ADAS algorithm verification method based on carra platform implements the steps of the ADAS algorithm verification method based on carra platform according to any one of claims 1 to 7 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110041907.2A CN112859810A (en) | 2021-01-13 | 2021-01-13 | ADAS algorithm verification method and device based on Carla platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110041907.2A CN112859810A (en) | 2021-01-13 | 2021-01-13 | ADAS algorithm verification method and device based on Carla platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112859810A true CN112859810A (en) | 2021-05-28 |
Family
ID=76003269
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110041907.2A Pending CN112859810A (en) | 2021-01-13 | 2021-01-13 | ADAS algorithm verification method and device based on Carla platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112859810A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902430A (en) * | 2019-03-13 | 2019-06-18 | 上海车右智能科技有限公司 | Traffic scene generation method, device, system, computer equipment and storage medium |
| CN110059393A (en) * | 2019-04-11 | 2019-07-26 | 东软睿驰汽车技术（沈阳）有限公司 | A kind of emulation test method of vehicle, apparatus and system |
| CN110188683A (en) * | 2019-05-30 | 2019-08-30 | 北京理工大学 | A kind of automatic Pilot control method based on CNN-LSTM |
| CN110427827A (en) * | 2019-07-08 | 2019-11-08 | 辽宁工程技术大学 | It is a kind of it is multiple dimensioned perception and Global motion planning under autonomous driving network |
| CN110647839A (en) * | 2019-09-18 | 2020-01-03 | 深圳信息职业技术学院 | Method and device for generating automatic driving strategy and computer readable storage medium |
| CN110864913A (en) * | 2019-11-28 | 2020-03-06 | 苏州智加科技有限公司 | Vehicle testing method and device, computer equipment and storage medium |
| CN110930811A (en) * | 2019-11-11 | 2020-03-27 | 北京交通大学 | System suitable for unmanned decision learning and training |
| CN111814308A (en) * | 2020-06-08 | 2020-10-23 | 同济大学 | Acceleration test system for automatic driving system |
| CN111841012A (en) * | 2020-06-23 | 2020-10-30 | 北京航空航天大学 | Automatic driving simulation system and test resource library construction method thereof |
| CN111859674A (en) * | 2020-07-23 | 2020-10-30 | 深圳慕智科技有限公司 | Automatic driving test image scene construction method based on semantics |
| CN112100856A (en) * | 2020-09-17 | 2020-12-18 | 上汽大众汽车有限公司 | Automatic driving joint simulation method based on multiple platforms |
-
2021
- 2021-01-13 CN CN202110041907.2A patent/CN112859810A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902430A (en) * | 2019-03-13 | 2019-06-18 | 上海车右智能科技有限公司 | Traffic scene generation method, device, system, computer equipment and storage medium |
| CN110059393A (en) * | 2019-04-11 | 2019-07-26 | 东软睿驰汽车技术（沈阳）有限公司 | A kind of emulation test method of vehicle, apparatus and system |
| CN110188683A (en) * | 2019-05-30 | 2019-08-30 | 北京理工大学 | A kind of automatic Pilot control method based on CNN-LSTM |
| CN110427827A (en) * | 2019-07-08 | 2019-11-08 | 辽宁工程技术大学 | It is a kind of it is multiple dimensioned perception and Global motion planning under autonomous driving network |
| CN110647839A (en) * | 2019-09-18 | 2020-01-03 | 深圳信息职业技术学院 | Method and device for generating automatic driving strategy and computer readable storage medium |
| CN110930811A (en) * | 2019-11-11 | 2020-03-27 | 北京交通大学 | System suitable for unmanned decision learning and training |
| CN110864913A (en) * | 2019-11-28 | 2020-03-06 | 苏州智加科技有限公司 | Vehicle testing method and device, computer equipment and storage medium |
| CN111814308A (en) * | 2020-06-08 | 2020-10-23 | 同济大学 | Acceleration test system for automatic driving system |
| CN111841012A (en) * | 2020-06-23 | 2020-10-30 | 北京航空航天大学 | Automatic driving simulation system and test resource library construction method thereof |
| CN111859674A (en) * | 2020-07-23 | 2020-10-30 | 深圳慕智科技有限公司 | Automatic driving test image scene construction method based on semantics |
| CN112100856A (en) * | 2020-09-17 | 2020-12-18 | 上汽大众汽车有限公司 | Automatic driving joint simulation method based on multiple platforms |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109557904B (en) | Test method, device, equipment and medium | |
| CN109413415B (en) | Camera controller testing system and testing method | |
| Bounini et al. | Autonomous vehicle and real time road lanes detection and tracking | |
| US10391406B2 (en) | Apparatus and method for safe drive inducing game | |
| EP3428840A1 (en) | Computer implemented detecting method, computer implemented learning method, detecting apparatus, learning apparatus, detecting system, and recording medium | |
| Reway et al. | Test methodology for vision-based adas algorithms with an automotive camera-in-the-loop | |
| KR102448358B1 (en) | Camera evaluation technologies for autonomous vehicles | |
| Dueholm et al. | Trajectories and maneuvers of surrounding vehicles with panoramic camera arrays | |
| CN111462249B (en) | Traffic camera calibration method and device | |
| Galko et al. | Vehicle-Hardware-In-The-Loop system for ADAS prototyping and validation | |
| CN110609288A (en) | Performance test method and device of automatic parking system | |
| EP4339938A1 (en) | Projection method and apparatus, and vehicle and ar-hud | |
| US20130155190A1 (en) | Driving assistance device and method | |
| US11250279B2 (en) | Generative adversarial network models for small roadway object detection | |
| CN112326258B (en) | Method, device and system for detecting automatic driving state and electronic equipment | |
| CN115408779A (en) | Simulation test method, device and storage medium for passenger-riding parking algorithm | |
| CN112562406B (en) | Method and device for identifying off-line driving | |
| CN114913506A (en) | 3D target detection method and device based on multi-view fusion | |
| KR20210086583A (en) | Method and apparatus for controlling driverless vehicle and electronic device | |
| JP5825713B2 (en) | Dangerous scene reproduction device for vehicles | |
| KR102095333B1 (en) | Augmented reality based vehicle test method and system | |
| CN114779665A (en) | Automatic parking simulation test method and device and readable storage medium | |
| Genevois et al. | Augmented reality on lidar data: Going beyond vehicle-in-the-loop for automotive software validation | |
| CN112859810A (en) | ADAS algorithm verification method and device based on Carla platform | |
| US20220315033A1 (en) | Apparatus and method for providing extended function to vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210528 |
|
| RJ01 | Rejection of invention patent application after publication |