WO2022083409A1 - Detection method and simulation method for abnormal road surface of road, and related apparatus - Google Patents

Abstract

The present application relates to a method and apparatus for detecting an abnormal road surface of a road, and a simulation method, a vehicle, and a medium. According to the correlations between recognized abnormal road surfaces of a target road and a target abnormal road surface corresponding to the target road, unrecognized abnormal road surfaces which may probably exist but have not been recognized on the target road can be predicted. The predicted unrecognized abnormal road surfaces can effectively compensate for the lack of recognition of abnormal road surfaces caused by the limitations of the layout and performance of a roadside sensing apparatus and the simulation performance of a simulation platform. Comprehensively recognized abnormal road surfaces and the predicted unrecognized abnormal road surfaces can be closer to actual abnormal road surfaces of the target road. Therefore, the realness of simulation of abnormal road surfaces of the target road can be effectively improved on the basis of the abnormal road surfaces and the predicted unrecognized abnormal road surfaces, thereby improving the realness of the simulation of a real road scene.

Description

Detection method, simulation method and related device for abnormal road surface

This application claims the priority of the Chinese patent application filed on October 24, 2020 with the application number 202011150953.8 and the application title is "Detection Method, Device, Simulation Method, Vehicle and Medium for Road Abnormal Pavement", all of which are The contents are incorporated herein by reference.

technical field

The present application relates to the field of computer vision technology, and in particular, to the detection and simulation of abnormal road surfaces.

Background technique

Since the simulation test can simulate various real road scenarios, it solves the problem of incomplete testing due to limited test sites during real vehicle testing. In the development process of vehicles, especially intelligent driving vehicles, simulation testing will be introduced into the entire development process. In order to realize the simulation test of various scenarios. When performing a simulation test, it is necessary to obtain real road scene information, simulate a road virtual scene based on the information, and then perform a simulation test on the vehicle on the virtual scene.

In the related art, the real scene of the road is sensed by the roadside sensing device, and it is determined whether there is an abnormal road surface thereon, and if there is, a simulation image of the abnormal road surface is displayed.

SUMMARY OF THE INVENTION

Based on this, it is necessary to provide a detection method, a simulation method and a related device for a road abnormal road surface that can solve the problem of the lack of abnormal road surface identification in order to solve the above technical problems.

On the one hand, an embodiment of the present application provides a method for detecting an abnormal road surface, comprising the following steps:

Obtain the real pavement information of the target road;

Identify the abnormal road surface on the target road according to the real road surface information;

According to the correlation between the identified abnormal road surface and the target abnormal road surface corresponding to the target road, the unidentified abnormal road surface on the target road is determined.

On the other hand, an embodiment of the present application provides a device for detecting abnormal road surfaces, including:

The acquisition module is used to acquire the real road surface information of the target road;

The identification module is used to identify the abnormal road surface on the target road according to the real road surface information;

The determining module is configured to determine the unidentified abnormal road surface on the target road according to the correlation between the recognized abnormal road surface and the target abnormal road surface corresponding to the target road.

On the other hand, an embodiment of the present application provides a method for simulating abnormal road surfaces, comprising the following steps:

Obtain the real pavement information of the target road;

Identify the abnormal road surface on the target road according to the real road surface information;

Determine the unidentified abnormal road surface on the target road according to the correlation between the recognized abnormal road surface and the target abnormal road surface corresponding to the target road;

The abnormal road surface and the determined unidentified abnormal road surface are simulated and displayed.

A computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above detection method when the computer program is executed.

In another aspect, an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the detection method or the simulation method in the foregoing aspect is implemented.

In another aspect, an embodiment of the present application provides a computer program product including instructions, which, when run on a computer, causes the computer to execute the detection method or the simulation method in the above aspect.

It can be seen that, according to the correlation between the identified abnormal road surface in the target road and the target abnormal road surface corresponding to the target road, the unidentified abnormal road surface that may exist on the target road but has not been identified can be estimated. The unrecognized abnormal road surface can effectively make up for the lack of abnormal road recognition caused by the layout and performance of the roadside sensing device and the simulation performance of the simulation platform. It can be closer to the actual abnormal road surface of the target road, so that the abnormal road surface based on the abnormal road surface and the estimated unidentified abnormal road surface can effectively improve the authenticity of the abnormal road surface simulation of the target road, thereby improving the authenticity of the real road scene simulation.

Description of drawings

1 is an application scenario diagram of a method for detecting abnormal road surfaces in an embodiment;

2 is a system architecture diagram of a method for detecting abnormal road surfaces of roads in one embodiment;

3 is a schematic flowchart of a method for detecting an abnormal road surface of a road in one embodiment;

4 is a schematic flowchart of step S306 in the first embodiment;

FIG. 5 is a flow chart of obtaining correlation coefficients between abnormal road surfaces in one embodiment;

6 is a schematic flowchart of step S404 in the second embodiment;

7 is a schematic flowchart of step S602 in one embodiment;

8 is a schematic flowchart of step S606 in one embodiment;

9 is a structural block diagram of an apparatus for detecting abnormal road surfaces of roads in one embodiment;

FIG. 10 is a flowchart of a method for simulating an abnormal road surface of a road in one embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.

The method for detecting abnormal road surfaces provided by the present application can be applied to the application scenario shown in FIG. 1 , which is an automatic driving simulation scenario, in which the simulation system simulates the real abnormal road surface on the target road for the automatic driving simulator to carry out Vehicle performance simulation test.

Based on the above application scenarios, the method for detecting abnormal road surfaces provided by the present application can be implemented through the system architecture shown in FIG. 2 . The roadside sensing device 102, the computer equipment 104 on which the simulation platform is deployed, and the transmission channel 106 constitute a simulation system, and the simulation system may be a digital twin simulation system. The so-called digital twin refers to making full use of physical model, sensor update, operation history and other data, integrating multi-disciplinary, multi-physics, multi-scale, multi-probability simulation process, and completing the mapping in virtual space, thus reflecting the whole of the corresponding physical equipment. Life cycle process, digital twin simulation system is a simulation system built based on digital twin technology.

In this system, the roadside sensing device 102 collects the real pavement information of the target road, and sends it to the computer device 104 through the transmission channel 106. The computer device 104 acquires the real pavement information of the target road, and identifies the target according to the real pavement information. The abnormal road surface on the road, and when the abnormal road surface is recognized, the current unrecognized abnormal road surface on the target road is determined according to the correlation between the abnormal road surface and the currently unrecognized abnormal road surface on the target road.

Wherein, the roadside sensing device 102 may be one or more roadside cameras disposed on one or both sides of the target road; the computer device 104 builds a simulation platform through the installed simulator or simulation software. The computer device may be a terminal device such as a personal computer, a notebook computer, a tablet computer, etc., or a server. The server may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a server. Cloud servers that provide cloud computing services.

The transmission channel 106 may be a network channel such as 3G, 4G, and 5G. In addition, the vehicle 108 can also obtain the road information in front of the vehicle through the in-vehicle camera device, and send the information to the simulation platform 104 through the transmission channel 106 to determine the accuracy of the abnormal road surface obtained by the computer device 104 through the above detection method. In addition, the computer device 104 can also send the abnormal road surface obtained by the above detection method to the vehicle 108 through the transmission channel 106, so that the vehicle 108 can give an early warning according to the abnormal road surface, so that the driver or the unmanned vehicle can take corresponding measures.

The implementation details of the technical solutions of the embodiments of the present application are described in detail below.

In an embodiment of the present application, as shown in FIG. 3 , a method for detecting abnormal road surface of a road is provided. Taking the method applied to the computer device in FIG. 2 as an example, the method for detecting abnormal road surface of a road can be described as follows: Include the following steps:

S302, obtain the real road surface information of the target road.

The target road refers to a real road that needs to detect abnormal road surfaces on the road, and may include expressways, first-class roads, second-class roads, and third-class roads, etc., which are not limited here. Among them, the expressway has the highest level, and its main function is for vehicles to drive at high speed; the first-class highway is located under the expressway, and its main function is to connect major economic and political centers; the second-class highway is located under the first-class highway, and its main function is to serve the Specific functional areas; the third-class highway is located under the second-class highway, and its main function is to connect local counties and towns.

The real road surface refers to the road surface on the real road, such as the real road surface on the highway. The real road surface information refers to information that can reflect the real road surface, which may be image information, video information, data information, and the like, and may be specifically obtained through the roadside sensing device 102 . For example, the real road surface information can be acquired through the roadside sensing device 102 such as a roadside camera, and the real road surface information is image information, and then the real road surface information is sent to the computer device 104 through the transmission channel 106, and the computer device 104 receives the real road surface information. information.

S304, according to the real road surface information, identify the abnormal road surface on the target road.

Abnormal road surface refers to the abnormal situation on the road surface that affects normal driving, which corresponds to the normal road surface, and may include the road surface with abnormal conditions such as pits, bulges, damage to speed bumps, road diseases, and missing manhole covers.

After obtaining the real road surface information of the target road, the computer device 104 may identify whether there is an abnormal road surface on the target road according to the real road surface information and according to a preset analysis strategy. For example, when the real road surface information is image information, the image information can be processed to obtain the characteristic information of the road surface in the image, and compared with the pre-stored characteristic information of the abnormal road surface to determine whether there is abnormal road surface in the image, If it exists, it means that there is an abnormal road surface on the target road; otherwise, it means that there is no abnormal road surface on the target road.

S306, according to the correlation between the identified abnormal road surface and the target abnormal road surface corresponding to the target road, determine the unidentified abnormal road surface on the target road.

It is understandable that, in actual roads, if the quality of a road is good, there will be fewer abnormal road surfaces on the road, and on the contrary, there will be more abnormal road surfaces. If the quality is better, the possibility of abnormal situation is less, and there will be less abnormal road surface, and the quality of the third-class road is poor, then the possibility of abnormal situation is higher, and the existing abnormal road surface will be more. Therefore, if there is one abnormal road surface on a certain road, the possibility of another abnormal road surface will be higher, that is, there is a certain correlation between abnormal road surfaces on the same type of road. After obtaining an abnormal road surface, the possibility of another abnormal road surface can be inferred. Therefore, when the computer device 104 recognizes that there is an abnormal road surface on the target road, it can infer whether there is a currently unrecognized road on the target road according to the abnormal road surface and the correlation between the abnormal road surface and the target abnormal road surface corresponding to the target road. Abnormal road surface not identified.

Among them, the same type of road refers to the same type of road, specifically refers to the road with similar pavement quality, for example, all expressways are the same type of road, all primary roads are the same type of road, all secondary roads are the same type of road, etc. Correlation refers to the degree of correlation between two variables. The correlation between abnormal road surfaces refers to the degree of correlation between the same or different abnormal road surfaces on the same type of road. The abnormal road surface identified according to the real road surface information and the target abnormal road surface are related. The correlation refers to the degree of correlation between the identified abnormal road surface and the target abnormal road surface. Unidentified abnormal road surface refers to the abnormal road surface on the target road that may not have been identified except the abnormal road surface identified according to the real road surface information. Here, it refers to the abnormal road surface that needs to be determined based on speculation. The types of abnormal road surfaces can be the same or different. For example, the identified abnormal road surface and the determined unidentified abnormal road surface are both road surfaces with pits, or the recognized abnormal road surface is a road surface with pits, and the determined unrecognized abnormal road surface is a road surface with pits. Identify the abnormal road surface as a road surface with bumps.

In this example, the currently unrecognized abnormal road surface that may exist on the target road can be obtained by inference based on the abnormal road surface that has been identified and the correlation between the abnormal road surface and the currently unrecognized abnormal road surface. The sensing device 102 is abnormal (such as failure), the transmission channel 106 is abnormal (such as network instability leads to data loss), and the simulation performance of the computer equipment 104 is insufficient (such as poor information restoration capability), and the roadside sensing device 102 is not fully covered (such as only Some roads are equipped with roadside sensing device 102) or the roadside sensing device 102 is blocked (such as blocked by obstacles) and other uncontrollable factors bring about the lack of recognition of abnormal road surfaces, which in turn leads to the problem of lack of simulation of abnormal road surfaces. The abnormal road surface can truly reflect the actual abnormal road conditions.

Further, the computer device 104 can also simulate and display the identified abnormal road surface and the presumably obtained unrecognized abnormal road surface. The virtual image corresponding to the unidentified abnormal road surface is displayed, and then the virtual image is simulated and displayed. It should be noted that all abnormal road surfaces encountered by real vehicles can be stored in the database of the simulation system, and virtual images are generated. After the computer device 104 determines all abnormal road surfaces on the target road, the corresponding abnormal roads can be retrieved from the database. The virtual image of the road surface is simulated and displayed.

In this embodiment, by estimating the unrecognized abnormal road surface according to the correlation between the recognized abnormal road surface and the target abnormal road surface, it can effectively solve the problems caused by the layout and performance of the roadside sensing device and the simulation performance of the computer equipment. Restrictions, etc., resulting in the lack of identification of abnormal road surfaces, which in turn leads to the problem of lack of abnormal road surface simulation, thereby effectively improving the authenticity of abnormal road surface simulation, thereby improving the authenticity of road real scene simulation.

In an embodiment of the present application, referring to FIG. 4 , S306 includes:

S402: Determine a target correlation coefficient between the abnormal road surface and the target abnormal road surface.

The correlation coefficient is a measure of the degree of linear correlation between variables. The simple correlation coefficient is also called the correlation coefficient, which is used to measure the linear relationship between two variables. The correlation coefficient between abnormal roads refers to the same or different road types on the same type of road. The correlation coefficient between abnormal road surfaces, the correlation coefficient can be preset.

In an embodiment of the present application, referring to FIG. 5 , S402 includes:

S502: Obtain the road surface type of the abnormal road surface existing on the target road or on the road with the same type as the target road.

S504, according to the road type of the abnormal road surface and the road type of the target road, obtain the target correlation coefficient between the abnormal road surface and the target abnormal road surface from a preset correlation coefficient between abnormal road surfaces.

Among them, the preset abnormal road surface is the abnormal road surface that may appear on all road types. According to the road type of the target road, the target abnormal road surface related to the road type of the target road can be selected from the preset abnormal road surface. Abnormal road surfaces that may appear under the road type of the road. According to the road surface type of the abnormal road surface and the target abnormal road surface, the target correlation coefficient between the abnormal road surface and the target abnormal road surface can be determined from the correlation coefficient.

After determining the abnormal road surface according to the real road surface information, the computer device 104 can further determine the road surface type of the abnormal road surface according to the characteristic information of the abnormal road surface, and can also determine the road type of the target road according to the user input parameters or the characteristic information of the target road. . Then, the computer device 104 may search and obtain the correlation coefficient between the abnormal road surface and the target abnormal road surface from the database as the target correlation coefficient according to the road surface type of the abnormal road surface and the road type of the target road.

In a possible implementation manner, the embodiment of the present application provides a method for determining a relationship coefficient of a preset abnormal road surface, wherein the preset abnormal road surface includes a plurality of road surface types, and the method includes:

S01, acquiring the road surface type of the abnormal road surface existing on the road to be processed.

The road to be processed includes the target road or a road of the same type as the target road.

S02: Acquire the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within a preset time period.

In an embodiment of the present application, S02 includes: obtaining from a historical database the occurrence times of abnormal road surfaces corresponding to multiple road surface types when the vehicle travels on the road to be processed within a preset historical time period.

It should be noted that there are usually differences in the types of abnormal road surfaces that occur in different road types. For example, there are no manhole covers and speed bumps on highways, so there will be no abnormal road surfaces with lost manhole covers and speed bumps. Manhole covers, speed bumps, etc., so there are abnormal road surfaces where manhole covers are lost and speed bumps are damaged. Therefore, based on the number of times the vehicle travels on each type of abnormal road surface on the road to be processed within a preset period of time, the abnormal road surface is between each other. The correlation is more in line with the actual situation, which can make the estimation of the unidentified abnormal road more reliable.

Specifically, the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types encountered on the road to be processed can be acquired by the computer device 104 . For example, when the target road is a highway, the number of times of each type of abnormal road surface encountered by all real vehicles on the highway is obtained. There is no specific limitation here, as long as it is ensured that the abnormal road surfaces encountered by all real vehicles are in the same area with similar road quality.

Since abnormal road surfaces can easily lead to abnormal driving of vehicles, or even traffic accidents, in actual driving, real vehicles (there may be more than one vehicle driving on real roads) will automatically report the situation to an abnormal road surface every time they encounter an abnormal road surface. The road monitoring platform, such as the Internet of Vehicles cloud server, records the number of times of each type of abnormal road surface encountered by all real vehicles (the abnormal road surface may change dynamically due to road maintenance and other reasons, but this has no impact on this application). It is assumed that there are n types of abnormal road surfaces encountered by real vehicles, which are respectively called abnormal road 1, abnormal road 2, ..., abnormal road n. Further, the road monitoring platform counts the historical preset duration (historical preset duration). The number of times of abnormal road surface 1, abnormal road surface 2, ..., abnormal road surface n encountered by all real vehicles in the time span recorded by the road monitoring platform shall be determined according to the specific situation, and then the computer equipment 104 can obtain the data from the road monitoring platform. Obtain the number of times of abnormal road 1, abnormal road 2, ..., abnormal road n encountered by all real vehicles in the historical preset time period.

It should be noted that, in this step, the occurrence frequency of abnormal road surfaces is determined by the "historical encounters" of the vehicle. In fact, the "historical encounters" is only a way to obtain the occurrence frequency of abnormal road surfaces.

In another embodiment of the present application, S02 includes:

According to the vehicle driving route, the number of abnormal road surfaces corresponding to each road surface type is estimated when the vehicle travels on the road to be processed within a preset time period in the future.

For example, when a roadside perception device such as a roadside camera captures an abnormal road on the target road or a road of the same type as the target road, and passes another roadside on the target road or on a road of the same type as the target road The sensing device, such as a roadside camera, detects that a certain vehicle will definitely pass through the road section where the abnormal road is located, then it can be determined that the vehicle will definitely encounter the abnormal road, and then through this method, it can be obtained that the vehicle will drive on the target road for a preset time in the future. The number of times each type of abnormal road surface was encountered on or on a road of the same type as the target road. In this way, the frequency of occurrence of abnormal road surfaces can be obtained by means of prediction based on the driving route of the vehicle.

It should be noted that, in this step, the frequency of occurrence of abnormal road surfaces is determined by the "driving route" of the vehicle. In fact, the "driving route" is only a way to obtain the frequency of occurrence of abnormal road surfaces. In practical applications, other methods may also be used to obtain the frequency of occurrence of abnormal road surfaces, which is not limited here.

S03: Determine the correlation coefficient between the preset abnormal road surfaces according to the number of occurrences of the abnormal road surfaces.

In an embodiment of the present application, the plurality of road surface types include a first abnormality type and a second abnormality type, and S03 includes:

The preset duration is divided into intervals to obtain a plurality of sub durations; the first occurrences of the abnormal road surface of the first abnormal type and the abnormal road surface of the second abnormal type in each sub duration are obtained. the second occurrence number; obtain the first mean square error of the first occurrence number and the second mean square deviation of the second occurrence number, and the covariance between the first occurrence number and the second occurrence number; According to the first mean square error, the second mean square error and the covariance, a correlation coefficient between the abnormal road surface of the first abnormal type and the abnormal road surface of the first abnormal type is determined.

For example, after obtaining the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within the preset period of time through the above method, the computer device 104 may divide the preset period of time into m sub-sections on average. Duration (m is greater than or equal to 2, there is no specific limit here), and record the number of occurrences of abnormal road 1, abnormal road 2, ... and abnormal road n encountered by all real vehicles within the t-th sub-duration, respectively x _{1, t} , x _2,t ,..., x _n,t , that is, in the t-th sub-duration, the number of occurrences of all real vehicles encountering abnormal road 1 is x _1,t , and the number of occurrences of encountering abnormal road 2 is x _{2 , t} , ..., the number of occurrences of encountering abnormal road n is x _n,t .

Then, the correlation coefficient between any two different abnormal road surfaces, ie, any two types of abnormal road surfaces, can be determined based on probability theory and quantitative statistics, using the times obtained in the preceding steps. Assuming that the correlation coefficient between the abnormal road i and the abnormal road j needs to be obtained at present, the number of times that all vehicles encounter the abnormal road i in the entire preset time period can be obtained by the aforementioned method, so as to obtain a plurality of first occurrence times x _i,1 , _xi,2 , . . . , _xi,m , and the number of times that all vehicles encounter abnormal road j, to obtain a plurality of second times as x _j,1 , x _j,2 , . . . , x _{j ,m} , then calculate the mean square error of multiple first occurrences as the first mean square error

and calculate the mean square error of multiple second occurrences as the second mean square error

as well as

Calculate the covariance between multiple first occurrences and multiple second occurrences

Finally, according to the first mean square error, the second mean square error and the covariance, the correlation coefficient between the abnormal road surface i of the first abnormal type and the abnormal road surface j of the second abnormal type is determined by the following methods:

In the formula, _ci,j represents the correlation coefficient between the abnormal road i and the abnormal road j, and the correlation coefficient satisfies symmetry, that is _{, ci,j} =c _j,i , m represents the number of sub-periods, and t represents the number of sub-periods. t sub-duration, x _i,t represents the number of times of abnormal road i encountered by all real vehicles in the t-th sub-duration, x _{j, t} represents the number of abnormal road j encountered by all real vehicles in the t-th sub-duration. It should be noted that, for the correlation coefficient between two identical abnormal road surfaces, that is, the same abnormal road surface, its value is 1, for example, c _i,i =1.

It should be noted that the acquisition process of the correlation coefficient between other abnormal road surfaces is the same as the acquisition process of the correlation coefficient between the abnormal road surface i and the abnormal road surface j, and will not be repeated here.

After obtaining the correlation coefficient between the preset abnormal road surfaces through the above method, the correlation coefficient, the road surface type corresponding to the correlation coefficient, and the road type corresponding to the correlation coefficient are stored in the database of the computer device 104 correspondingly, In actual use, the correlation coefficient is called by the computer device 104, and the target correlation coefficient between the abnormal road surface and the target abnormal road surface is determined according to the correlation coefficient.

S404, according to the target correlation coefficient, determine the unidentified abnormal road surface on the target road.

Assuming that the abnormal road surface obtained by the computer device 104 according to the real road surface information of the target road is abnormal road surface i, then when all possible abnormal road surfaces on the target road and roads of the same type as the target road include the above abnormal road surface 1 and abnormal road surface 2 , ..., abnormal road surface n, the obtained first correlation coefficients between abnormal road surface i and abnormal road surface 1, 2, ..., n are c _1,i , c _2,i , ..., c _{n , respectively, i} . Then, the computer device 104 sorts the obtained first correlation coefficients c1 _,i , c2 _{, i} , . pavement.

In the above embodiment, the unrecognized abnormal road surface can be estimated according to the road surface type of the existing abnormal road surface, the road type of the target road and the preset abnormal road surface, and the preset abnormal road surface can be estimated. The correlation coefficient between them is obtained based on the number of occurrences of abnormal road surfaces of each type of abnormal road surfaces encountered by vehicles traveling on the road to be processed within a preset period of time, which corresponds to the target road, thus making the estimated abnormal road surface. more reliable.

In an embodiment of the present application, referring to FIG. 6 , S404 includes:

S602, obtaining the true rate of the abnormal road surface.

The true rate of the abnormal road surface refers to the probability that the abnormal road surface determined according to the real road surface information of the target road can reflect the real road surface on the target road. Specifically, the abnormal road surface determined by the computer device 104 according to the real road surface information can reflect the actual road surface on the target road. The probability of the road surface, which is affected by various factors, such as the acquisition performance of the roadside sensing device 102, the transmission performance of the transmission channel 106, and the simulation performance of the computer equipment 104, so the real abnormal road surface can be determined based on these influence information. Rate.

In an embodiment of the present application, referring to FIG. 7 , the acquisition of the true rate of the abnormal road surface in S602 includes:

S702 , acquiring the collection truth rate, transmission missing rate and restoration success rate of the real road surface information.

The collection truth rate refers to the probability that the collected real road surface information is real. Specifically, it may be the probability that the real road surface information captured by the roadside perception device 102 such as the roadside camera is real, which is an attribute parameter of the roadside perception device 102 and can be Obtained from the specification of the roadside sensing device 102 or through experimental tests in advance. In practical applications, the failure rate of the roadside sensing device 102 can be used as the actual rate of collection.

The transmission missing rate refers to the missing rate of the real road surface information in the transmission process, which can be specifically the process of the roadside sensing device 102 such as the roadside camera transmitting the captured real road surface information to the computer equipment 104 through the transmission channel 106 such as the network. The probability of missing real road surface information, which is an attribute parameter of the transmission channel 106, can be obtained through experimental tests in advance. In practical applications, the packet loss rate of the transmission channel 106 can be used as the transmission missing rate.

The restoration success rate refers to the success rate of restoring the received real road information, specifically the success rate of the computer device 104 restoring the received real road information, which is an attribute parameter of the computer device 104 and can be obtained from the description document of the computer device 104 Or obtained in advance through experimental tests.

S704: Acquire the true rate of the abnormal road surface according to one or more of the acquisition true rate, the transmission missing rate, and the restoration success rate.

It can be seen from the foregoing that the acquisition real rate corresponds to the roadside sensing device 102 , the transmission missing rate corresponds to the transmission channel 106 , and the restoration success rate corresponds to the computer equipment 104 , and the roadside sensing device 102 , the transmission channel 106 and the computer equipment 104 are three related and independent The device can be considered to be related and independent during operation. Therefore, after obtaining the real road information collection truth rate, transmission missing rate and restoration success rate, it can be based on one or more of the collection truth rate, transmission missing rate and restoration success rate. It uses the method of probability and statistics to calculate and obtain the true rate of abnormal road surface. Assuming that the acquired real rate of acquisition is p _sensor , the transmission missing rate is p _transmission , and the restoration success rate is p _refresh , then the calculated real rate of abnormal road surface can be p _real =(1-p _sensor )(1-p _transmission )(1-p _refresh ). Understandably, the true rate is ideally 1.

It should be noted that the true rate of abnormal road surface can be obtained in advance through the above method, and then stored in the database of the computer device 104, and can be called directly when used; The truth rate, transmission missing rate and restoration success rate are stored in the database of the computer device 104, and one or more of them can be selected according to actual needs to calculate and obtain the truth rate of abnormal road surfaces.

Step 604: Determine the existence rate of the target abnormal road surface according to the true rate of the abnormal road surface and the target correlation coefficient.

The existence rate of the target abnormal road surface refers to the probability that any type of abnormal road surface that may exist on the target road or a road of the same type as the target road exists on the target road.

When determining that there is an abnormal road surface i on the target road according to the real information of the road surface, the computer device 104 will determine that the true rate of the abnormal road surface i is p _reali . If the correlation between abnormal road j and abnormal road i is large, then the probability of abnormal road j on the target road is high, otherwise it is small, which means that the probability of whether the target road has abnormal road j is proportional to its and abnormal road i Therefore, when the abnormal road surface i is obtained, the probability that the abnormal road surface j exists on the target road, that is, the existence rate is p _{realic i,j} , and so on, the above abnormal road surface 1, abnormal road surface 2, ..., the probability of the abnormal road surface n, that is, the existence rate are respectively p _{realic 1,i} , p _{realic 2,i} , ..., p _{realic n,i} , that is, the existence rate of the currently unrecognized abnormal road surface is respectively p _{realic 1 ,i} , p _{realic 2,i} , …, p _{realic n,i} . It should be noted that since c _i,i =1, p _reali c _i,i =p _reali .

S606, according to the existence rate of the target abnormal road surface, determine the unrecognized abnormal road surface on the target road.

After obtaining the existence rate of the target abnormal road surface, the computer device 104 can determine whether there is an unrecognized abnormal road surface on the target road according to the existence rate. For example, the existence rate of abnormal road surface 1, abnormal road surface 2, . Whether the existence rate of the above abnormal road surface 1, abnormal road surface 2, ..., abnormal road surface n is higher than the preset existence rate, if it is higher, it is considered that the corresponding abnormal road surface exists on the target road. Of course, other methods may also be used to determine whether there is a currently unrecognized abnormal road surface on the target road according to the existence rate of the currently unrecognized abnormal road surface.

In an embodiment of the present application, referring to FIG. 8 , in S606, according to the existence rate of the target abnormal road surface, the unrecognized abnormal road surface on the target road is determined, including:

S802 , obtaining the random rate of the target abnormal road surface, wherein the random rate is determined according to a random number, and the random number obeys a uniform distribution of 0-1.

The random rate refers to the probability that the target abnormal road surface randomly appears, which can be represented by random numbers. For the abnormal road surface 1, abnormal road surface 2, ..., abnormal road surface n, the computer device 104 will generate n random numbers obeying a 0-1 uniform distribution, which are respectively denoted as ε ₁ , ε ₂ , ..., ε _n . Among them, the generation method of random numbers can adopt existing tools, such as simulation software MATLAB.

S804, according to the existence rate and random rate of the target abnormal road surface, determine the unidentified abnormal road surface on the target road.

According to an embodiment of the present application, the current unrecognized abnormal road surface on the target road is determined according to the existence rate and random rate of the currently unrecognized abnormal road surface, including:

determining an undetermined road surface whose random rate is less than the existence rate from the target abnormal road surface;

The unidentified abnormal road surface on the target road is determined according to the undetermined road surface.

After obtaining the random numbers ε ₁ , ε ₂ , . . . , ε _n , the computer device 104 can compare these random numbers with the corresponding existence rates of abnormal road surfaces to determine whether there are currently unrecognized abnormal road surfaces on the target road. For example, it is successively judged whether ε ₁ ≤p _{realic 1,i} , ε ₂ ≤p _{realic 2,i} , . . . , ε _n ≤p _{realic n,i} are established, and if so, the corresponding abnormal road surface is taken as the target road If ε _k ≤ p _{realic k,i} is established, it is determined that there will be abnormal road k on the target road.

Therefore, on the premise that the abnormal road surface on the target road is determined according to the real road surface information of the target road, the unidentified abnormal road surface that may exist on the target road can be effectively determined according to the correlation between the abnormal road surface and the target abnormal road surface. , that is, according to the abnormal road surface that has been photographed, the abnormal road surface that has not been photographed can be estimated or predicted, which can assist the simulation system to comprehensively determine which abnormal road surfaces exist on the target road, and then compensate for the roadside sensing device, transmission channel or The lack of simulation performance of the computer equipment itself or the lack of road simulation caused by uncontrollable factors such as incomplete coverage of roadside sensing devices or occlusions effectively improves the authenticity of the simulation.

In order to verify whether the method for detecting abnormal road surface of the present application is effective, it can be tested in computer equipment, and the abnormal road surface that can be simulated by computer equipment at one time is counted. The statistical results are shown in Table 1:

Table 1

Compared with the real road, it is obvious that the performance of the present application is better than that of the related art, and the simulated abnormal road surface is more comprehensive.

It should be understood that although the steps in the flowcharts of FIGS. 3-8 are shown in sequence according to the arrows, these steps are not necessarily executed in the sequence shown by the arrows. Unless explicitly stated herein, there is no strict order limit to the execution of these steps, and the steps may be performed in a currently unidentified order. Moreover, at least a part of the steps in Figs. 3-8 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but may be executed at different times. The order of execution is also not necessarily sequential, but may be performed alternately or alternately with the currently unidentified step or at least a portion of the sub-steps or phases of the currently unidentified step.

To sum up, according to the method for detecting abnormal road surface of the road according to the embodiment of the present application, by estimating the currently unrecognized abnormal road surface according to the abnormal road surface that has been identified and the correlation between the abnormal road surfaces, it is possible to effectively solve the problem caused by roadside perception. The layout and performance of the device and the limitation of the simulation performance of the computer equipment, etc., lead to the problem of the lack of abnormal road simulation, which effectively improves the authenticity of the abnormal road simulation, thereby improving the authenticity of the real road scene simulation.

It should be noted that the method for detecting abnormal road surfaces of the present application can be applied to simulation tests, such as in an automatic driving simulation system, where the simulation system simulates the real abnormal road surfaces on the road for testing by the automatic driving simulator. It can be applied to vehicle control, such as predicting potential abnormal road surfaces on the road to provide a safe driving reference for drivers or autonomous vehicles.

In one embodiment, a device for detecting abnormal road surface of a road is provided. Referring to FIG. 9 , the device 900 for detecting abnormal road surface of a road may include: an acquisition module 902 , an identification module 904 and a determination module 906 .

Among them, the acquisition module 902 is used to acquire the real road surface information of the target road; the identification module 904 is used to identify the abnormal road surface on the target road according to the real road surface information; the determination module 906 is used to identify the target road corresponding to the target road according to the abnormal road surface. Correlation between abnormal pavements to determine unidentified abnormal pavements on the target road.

In one embodiment, the determining module 906 is specifically configured to determine the target correlation coefficient between the abnormal road surface and the target abnormal road surface; and determine the unidentified abnormal road surface on the target road according to the target correlation coefficient.

Further, the determination module 906 is specifically used to obtain the road type of the abnormal road surface and the road type of the target road; according to the road type of the abnormal road surface and the road type of the target road, obtain the abnormal road surface and the abnormal road surface from the correlation coefficient between the preset abnormal road surfaces. Target correlation coefficient between target abnormal road surfaces.

In another embodiment, the determining module 906 is specifically configured to obtain the truth rate of the abnormal road surface; determine the existence rate of the target abnormal road surface according to the truth rate of the abnormal road surface and the target correlation coefficient; determine the target according to the existence rate of the target abnormal road surface Unidentified abnormal pavement on the road.

Further, the determination module 906 is specifically used to obtain the collection truth rate, transmission loss rate and restoration success rate of the real road surface information; obtain the abnormal road surface information according to one or more of the collection truth rate, transmission loss rate and restoration success rate. true rate.

In yet another embodiment, the determining module 906 is specifically configured to obtain the random rate of the existence of the target abnormal road surface; and determine the unidentified abnormal road surface on the target road according to the existence rate and the random rate of the target abnormal road surface. Among them, the random rate is determined according to the random number, and the random number obeys the uniform distribution of 0-1.

Further, the determining module 906 is specifically configured to determine, from the target abnormal road surface, an undetermined road surface whose random rate is less than the existence rate; and determine the unidentified abnormal road surface on the target road according to the undetermined road surface.

For the specific definition of the device for detecting abnormal road surface, please refer to the above definition of the detection method for abnormal road surface, which will not be repeated here. Each module in the above-mentioned device for detecting abnormal road surface can be implemented in whole or in part by software, hardware and combinations thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a method for simulating abnormal road surfaces is provided. Referring to FIG. 10 , the method may include the following steps:

S1002, obtain the real road surface information of the target road.

S1004, according to the real road surface information, identify the abnormal road surface on the target road.

S1006, according to the correlation between the identified abnormal road surface and the target abnormal road surface corresponding to the target road, determine the unidentified abnormal road surface on the target road.

S1008, simulate and display the abnormal road surface and the unrecognized abnormal road surface.

In one embodiment, determining the unrecognized abnormal road surface on the target road according to the correlation between the recognized abnormal road surface and the target abnormal road surface corresponding to the target road includes:

determining a target correlation coefficient between the abnormal road surface and the target abnormal road surface;

The unidentified abnormal road surface on the target road is determined according to the target correlation coefficient.

In one embodiment, the determining a target correlation coefficient between the abnormal road surface and the target abnormal road surface includes:

obtaining the road type of the abnormal road surface and the road type of the target road;

According to the road surface type of the abnormal road surface and the road type of the target road, the target correlation coefficient between the abnormal road surface and the target abnormal road surface is obtained from a preset correlation coefficient between abnormal road surfaces.

In one embodiment, the preset abnormal road surface includes a plurality of road surface types, and the method further includes:

obtaining the road surface type of the abnormal road surface existing on the road to be processed, the road to be processed includes the target road or a road of the same type as the target road;

Acquire the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within a preset time period;

The correlation coefficient between the preset abnormal road surfaces is determined according to the number of occurrences of the abnormal road surfaces.

In one embodiment, the acquiring the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within a preset period of time includes:

Obtain the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within the preset period of time obtained from the history database; or,

The number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types, respectively, when the vehicle travels on the to-be-processed road within the preset time period in the future is estimated according to the vehicle driving route.

In one embodiment, the plurality of road surface types include a first abnormality type and a second abnormality type, and the determining the correlation coefficient between the preset abnormal road surfaces according to the number of occurrences of the abnormal road surface includes:

performing interval division on the preset duration to obtain a plurality of sub-durations;

Obtaining the first occurrence number of abnormal road surfaces of the first abnormal type and the second occurrence number of abnormal road surfaces of the second abnormal type when the vehicle travels on the road to be processed within each sub-duration;

obtaining the first mean square error of the first occurrence and the second mean square of the second occurrence, and the covariance between the first occurrence and the second occurrence;

According to the first mean square error, the second mean square error and the covariance, a correlation coefficient between the abnormal road surface of the first abnormal type and the abnormal road surface of the first abnormal type is determined.

In one embodiment, the determining the unidentified abnormal road surface on the target road according to the target correlation coefficient includes:

obtaining the true rate of the abnormal road surface;

determining the existence rate of the target abnormal road surface according to the true rate of the abnormal road surface and the target correlation coefficient;

The unrecognized abnormal road surface on the target road is determined according to the existence rate of the target abnormal road surface.

In one embodiment, the obtaining the true rate of the abnormal road surface includes:

Obtain the collection truth rate, transmission missing rate and restoration success rate of the real road information;

The true rate of the abnormal road surface is acquired according to one or more of the acquisition true rate, the transmission missing rate, and the restoration success rate.

In one embodiment, determining the unrecognized abnormal road surface on the target road according to the existence rate of the target abnormal road surface includes:

obtaining the random rate of the target abnormal road surface;

The unidentified abnormal road surface on the target road is determined according to the existence rate of the target abnormal road surface and the random rate.

In one embodiment, determining the unidentified abnormal road surface on the target road according to the existence rate of the target abnormal road surface and the random rate includes:

determining an undetermined road surface whose random rate is less than the existence rate from the target abnormal road surface;

The unidentified abnormal road surface on the target road is determined according to the undetermined road surface.

In one embodiment, the random rate is determined according to a random number, and the random number obeys a uniform distribution of 0-1.

In one embodiment, performing a simulation display on the abnormal road surface and the unrecognized abnormal road surface includes: acquiring a first virtual image corresponding to the abnormal road surface, and acquiring a second virtual image corresponding to the unrecognized abnormal road surface; and the second virtual image is displayed.

In one embodiment, a computer device is provided, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the processor implements the foregoing detection method for abnormal road surface or realizes the foregoing simulation method for abnormal road surface.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, implements a method for detecting abnormal road surfaces or a method for simulating abnormal road surfaces.

The embodiments of the present application also provide a computer program product including instructions, which, when executed on a computer, cause the computer to execute the methods provided by the above embodiments.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or currently unidentified medium used in the various embodiments provided in this application may include non-volatile and/or volatile memory. Nonvolatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Road (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments can be combined arbitrarily. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features It is considered to be the range described in this specification.

The above-mentioned embodiments only represent several embodiments of the present application, and the descriptions thereof are specific and detailed, but should not be construed as a limitation on the scope of the invention patent. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the scope of protection of the patent of the present application shall be subject to the appended claims.

Claims (16)

1. A method for detecting abnormal road surface of a road, the method is executed by a computer device, and the method comprises:

   Obtain the real pavement information of the target road;

   Identifying an abnormal road surface on the target road according to the real road surface information;

   According to the correlation between the identified abnormal road surface and the target abnormal road surface corresponding to the target road, the unidentified abnormal road surface on the target road is determined.
2. The method for detecting an abnormal road surface of a road according to claim 1, wherein the unrecognized abnormality on the target road is determined according to the correlation between the recognized abnormal road surface and the target abnormal road surface corresponding to the target road Pavement, including:

   determining a target correlation coefficient between the abnormal road surface and the target abnormal road surface;

   The unidentified abnormal road surface on the target road is determined according to the target correlation coefficient.
3. The method for detecting an abnormal road surface of a road according to claim 2, wherein the determining a target correlation coefficient between the abnormal road surface and the target abnormal road surface comprises:

   obtaining the road type of the abnormal road surface and the road type of the target road;

   According to the road surface type of the abnormal road surface and the road type of the target road, the target correlation coefficient between the abnormal road surface and the target abnormal road surface is obtained from a preset correlation coefficient between abnormal road surfaces.
4. The method for detecting an abnormal road surface according to claim 3, wherein the preset abnormal road surface includes a plurality of road surface types, and the method further comprises:

   obtaining the road surface type of the abnormal road surface existing on the road to be processed, the road to be processed includes the target road or a road of the same type as the target road;

   Acquire the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within a preset time period;

   The correlation coefficient between the preset abnormal road surfaces is determined according to the number of occurrences of the abnormal road surfaces.
5. The method for detecting an abnormal road surface of a road according to claim 4, wherein the obtaining the number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types when the vehicle travels on the road to be processed within a preset time period, comprising:

   The number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types, respectively, encountered by the vehicle traveling on the road to be processed within the preset period of time obtained from the history database; or,

   The number of occurrences of abnormal road surfaces corresponding to the plurality of road surface types, respectively, when the vehicle travels on the to-be-processed road within the preset time period in the future is estimated according to the vehicle driving route.
6. The method for detecting abnormal road surfaces of a road according to claim 4, wherein the plurality of road surface types includes a first abnormal type and a second abnormal type, and the number of occurrences of the abnormal road surfaces is determined to determine the difference between the preset abnormal road surfaces. The correlation coefficient includes:

   performing interval division on the preset duration to obtain a plurality of sub-durations;

   Obtaining the first occurrence number of abnormal road surfaces of the first abnormal type and the second occurrence number of abnormal road surfaces of the second abnormal type when the vehicle travels on the road to be processed within each sub-duration;

   obtaining the first mean square error of the first occurrence and the second mean square of the second occurrence, and the covariance between the first occurrence and the second occurrence;

   According to the first mean square error, the second mean square error and the covariance, a correlation coefficient between the abnormal road surface of the first abnormal type and the abnormal road surface of the first abnormal type is determined.
7. The method for detecting abnormal road surfaces according to any one of claims 2-6, wherein determining the unrecognized abnormal road surface on the target road according to the target correlation coefficient, comprising:

   obtaining the true rate of the abnormal road surface;

   determining the existence rate of the target abnormal road surface according to the true rate of the abnormal road surface and the target correlation coefficient;

   The unrecognized abnormal road surface on the target road is determined according to the existence rate of the target abnormal road surface.
8. The method for detecting an abnormal road surface of a road according to claim 7, wherein the obtaining the true rate of the abnormal road surface comprises:

   Obtain the collection truth rate, transmission missing rate and restoration success rate of the real road information;

   The true rate of the abnormal road surface is acquired according to one or more of the acquisition true rate, the transmission missing rate, and the restoration success rate.
9. The method for detecting abnormal road surfaces according to claim 7, wherein determining the unrecognized abnormal road surface on the target road according to the existence rate of the target abnormal road surface, comprising:

   obtaining the random rate of the target abnormal road surface;

   The unrecognized abnormal road surface on the target road is determined according to the existence rate of the target abnormal road surface and the random rate.
10. The method for detecting abnormal road surfaces according to claim 9, wherein determining the unidentified abnormal road surface on the target road according to the existence rate of the target abnormal road surface and the random rate, comprising:

    determining an undetermined road surface whose random rate is less than the existence rate from the target abnormal road surface;

    The unidentified abnormal road surface on the target road is determined according to the undetermined road surface.
11. The method for detecting abnormal road surfaces according to claim 9, wherein the random rate is determined according to a random number, and the random number obeys a uniform distribution of 0-1.
12. A detection device for abnormal road surface of a road, comprising:

    The acquisition module is used to acquire the real road surface information of the target road;

    an identification module, configured to identify the abnormal road surface on the target road according to the real road surface information;

    A determination module, configured to determine the unrecognized abnormal road surface on the target road according to the correlation between the abnormal road surface and the target abnormal road surface corresponding to the target road.
13. A method for simulating abnormal road surfaces, the method is executed by computer equipment, and the method comprises:

    Obtain the real pavement information of the target road;

    Identifying an abnormal road surface on the target road according to the real road surface information;

    According to the correlation between the identified abnormal road surface and the target abnormal road surface corresponding to the target road, determine the unidentified abnormal road surface on the target road;

    The abnormal road surface and the unrecognized abnormal road surface are simulated and displayed.
14. A computer device, comprising a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method according to any one of claims 1 to 11 when the processor executes the computer program, or implements when executing the computer program. The steps of the method of claim 13.
15. A computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the method of any one of claims 1 to 11, or, when executed, implements the method described in claim 13 steps of the method.
16. A computer program product comprising instructions which, when run on a computer, cause the computer to perform the steps of the method of any one of claims 1 to 11, or, when executed, implement the steps of the method of claim 13 .

Images