CN117681865A - Vehicle obstacle avoidance control system integrating road monitoring and evaluation method thereof - Google Patents
Vehicle obstacle avoidance control system integrating road monitoring and evaluation method thereof Download PDFInfo
- Publication number
- CN117681865A CN117681865A CN202311704926.4A CN202311704926A CN117681865A CN 117681865 A CN117681865 A CN 117681865A CN 202311704926 A CN202311704926 A CN 202311704926A CN 117681865 A CN117681865 A CN 117681865A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- information
- road
- obstacle avoidance
- module
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 33
- 238000011156 evaluation Methods 0.000 title claims abstract description 18
- 230000008447 perception Effects 0.000 claims abstract description 62
- 230000010365 information processing Effects 0.000 claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 21
- 238000012545 processing Methods 0.000 claims abstract description 10
- 230000004927 fusion Effects 0.000 claims description 58
- 238000012216 screening Methods 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 238000013139 quantization Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 abstract description 5
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 238000001514 detection method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 241001464837 Viridiplantae Species 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000006837 decompression Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013095 identification testing Methods 0.000 description 1
- 230000003116 impacting effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Abstract
A vehicle obstacle avoidance control system integrating road monitoring and an evaluation method thereof are provided, wherein the vehicle obstacle avoidance control system comprises: the road information acquisition module is at least used for acquiring road environment information; the road information processing module is at least used for processing the road environment information so as to obtain road obstacle avoidance information; the vehicle information acquisition module is at least used for acquiring vehicle environment information; the vehicle information processing module is at least used for generating vehicle obstacle avoidance information based on the vehicle environment information, and combining the vehicle obstacle avoidance information with the road obstacle avoidance information to assist and/or control the vehicle obstacle avoidance, so that the problem that in the prior art, effective technical means for expanding the perception range of a driver or an automatic driving vehicle are lacking, and risk problems existing in the vehicle using process cannot be estimated in advance is solved, and dangerous driving scenes caused by bad weather, green planting shielding and building shielding are avoided, so that the driving safety of a user is difficult to ensure.
Description
Technical Field
The invention belongs to the technical field of vehicle obstacle avoidance control, and particularly relates to a vehicle obstacle avoidance control system and a control method integrating road monitoring.
Background
With the development of automobile technology, the intelligent and networking degree of automobiles is higher and higher. In the prior art, an intelligent internet-connected vehicle uses a camera, a radar and an inertial navigation system to identify surrounding environment and obstacles so as to assist a user to drive safely. The road monitoring device monitors the corresponding road sections through the road monitoring lens in combination with other prior art means and assists the traffic management department to work. In the prior art, an effective technical means for expanding the perception range of a driver or an automatic driving vehicle is lacking, and the risk problem existing in the vehicle using process cannot be estimated in advance, so that dangerous driving scenes caused by bad weather, green planting shielding and building shielding are avoided, and the driving safety of a user is difficult to be ensured.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, the present invention aims to provide a vehicle obstacle avoidance control system integrating road monitoring and an evaluation method thereof, which solve the problems of bad weather, green plant shielding and dangerous driving scenes caused by building shielding, so that the driving safety of a user is difficult to be ensured.
To achieve the above object, a first aspect of the present application provides a vehicle obstacle avoidance control system that merges road monitoring, including:
the road information acquisition module is at least used for acquiring road environment information;
the road information processing module is at least used for processing the road environment information so as to obtain road obstacle avoidance information;
the vehicle information acquisition module is at least used for acquiring vehicle environment information;
and the vehicle information processing module is at least used for generating vehicle obstacle avoidance information based on the vehicle environment information, and combining the vehicle obstacle avoidance information and the road obstacle avoidance information to assist and/or control the vehicle so as to avoid the obstacle.
As one embodiment of the present invention, the road obstacle avoidance information includes a roadside data layer, a roadside feature layer, and a roadside target layer;
the vehicle obstacle avoidance information comprises a vehicle edge data layer, a vehicle edge feature layer and a vehicle edge target layer;
the fusing the vehicle obstacle avoidance information and the road obstacle avoidance information assists and/or controls a vehicle to avoid an obstacle, including: the vehicle information processing module assists and/or controls a vehicle to avoid an obstacle based on a front fusion of the roadside data layer and the roadside data layer, based on a rear fusion of the roadside target layer and the roadside target layer, and based on an intermediate fusion including the roadside feature layer or the roadside feature layer.
As an embodiment of the present invention, the vehicle obstacle avoidance control system includes a cloud end, where the cloud end is configured to store the obstacle avoidance information.
As one embodiment of the invention, the road information processing module comprises a road information sensing module and a road information screening module, wherein the road information sensing module is used for generating a first attribute of an obstacle based on the road environment information, and screening out road obstacle avoidance information from the road environment information based on the first attribute of the obstacle through the road information screening module.
As an implementation mode of the invention, the road information processing module further comprises a road information recording module and a road information uploading module, wherein the road information recording module is used for recording the road obstacle avoidance information, and the road information uploading module is used for uploading the recorded road obstacle avoidance information to a cloud.
As an implementation mode of the invention, the vehicle information processing module further comprises a vehicle information downloading module, a vehicle information recording module and a perception fusion module, wherein the vehicle information downloading module is used for downloading the road obstacle avoidance information, the vehicle information recording module is used for recording the road obstacle avoidance information, and the perception fusion module is used for fusing the vehicle obstacle avoidance information and the road obstacle avoidance information to generate an obstacle avoidance result.
As one embodiment of the invention, the vehicle information processing module comprises a vehicle information sensing module, a vehicle information screening module and an obstacle avoidance result generating module;
the vehicle information sensing module generates a second attribute of an obstacle and an obstacle confidence degree based on the vehicle environment information, and generates a vehicle blind area and vehicle obstacle avoidance information based on the second attribute of the obstacle and the obstacle confidence degree;
the vehicle information screening module is used for recording and transmitting road obstacle avoidance information from the vehicle information recording module and transmitting the road obstacle avoidance information to the perception fusion module;
the obstacle avoidance result generation module is used for avoiding the obstacle based on the obstacle avoidance result.
As one embodiment of the invention, the vehicle obstacle avoidance control system further comprises an alarm module, wherein the alarm module is used for giving an alarm before the vehicle information processing module controls the vehicle to avoid the obstacle.
A second aspect of the present application provides a vehicle obstacle avoidance control method that merges road monitoring, including:
collecting road environment information;
processing the road environment information to obtain road obstacle avoidance information;
acquiring vehicle environment information;
and generating vehicle obstacle avoidance information based on the vehicle environment information, and fusing the vehicle obstacle avoidance information and the road obstacle avoidance information to control the vehicle to avoid the obstacle.
A third aspect of the present application provides an evaluation method of a vehicle obstacle avoidance control system based on the fusion road monitoring according to the first aspect of the present application, including:
calculating a first perceived performance based on the vehicle obstacle avoidance information;
calculating a second perceived performance based on the vehicle obstacle avoidance information and the road obstacle avoidance information;
comparing the first perceived performance and the second perceived performance to obtain a verification result.
As one embodiment of the invention, the assessment method is tested in a bad weather scenario;
setting a sensor quantization level of a bad weather scene, wherein the sensor quantization level is used for measuring the influence degree of bad weather on a sensor;
and performing obstacle position and/or speed recognition tests on the same road section based on the sensor quantization level so as to compare the first perception performance with the second perception performance to obtain a verification result.
A fourth aspect of the present application provides an evaluation system of a vehicle obstacle avoidance control system based on the fusion road monitoring according to the first aspect of the present application, including:
the first perception calculation module is at least used for calculating first perception performance based on the vehicle obstacle avoidance information;
the second perception calculation module is at least used for calculating second perception performance based on the vehicle obstacle avoidance information and the road obstacle avoidance information; and the perception comparison module is at least used for comparing the first perception performance with the second perception performance to obtain a verification result.
A fifth aspect of the present application provides a readable storage medium storing a computer program which when executed by a processor performs the steps of the method according to any one of the second or third aspects of the present application.
In summary, compared with the prior art, the invention has at least one of the following beneficial technical effects:
1. the intelligent network-connected vehicle and road monitoring combined sensing capability is established, so that the sensing range of the vehicle is effectively enlarged, and the effect of active control or auxiliary control on the vehicle is realized;
2. through combining vehicle sensor information with road monitoring, effectively enlarge driver or the perception scope of autopilot vehicle, can realize the risk of assessing in advance, avoid because of weather reason, green plant shelters from or the danger that the building shelters from to ensure personnel and vehicle safety.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly introduced below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a block diagram of a prior art obstacle avoidance control system for a vehicle.
Fig. 2 is a schematic block diagram of a vehicle obstacle avoidance control system with road monitoring in an embodiment of the application.
Fig. 3 is a schematic block diagram of another vehicle obstacle avoidance control system incorporating road monitoring according to an embodiment of the present application.
FIG. 4 is a schematic diagram of a multi-modal fusion algorithm incorporating multiple fusion algorithms in one embodiment of the present application;
fig. 5 is a flow chart of a vehicle obstacle avoidance control method with road monitoring in an embodiment of the application.
Fig. 6 is a flow chart of an evaluation method of a vehicle obstacle avoidance control system or control method incorporating road monitoring in an embodiment of the present application.
FIG. 7 is a graph of an evaluation of perceptual performance in an embodiment of the present application.
Fig. 8 is a graph of road test performance results of the first perceived performance a according to an embodiment of the present application.
Fig. 9 is a graph of road test performance results of the second perceived performance B according to an embodiment of the present application.
Fig. 10 is a performance comparison graph of a first perceived performance a and a second perceived performance B according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, based on the embodiments herein, which are within the scope of the protection of the present application, will be within the skill of the art without inventive effort. Furthermore, it should be understood that the detailed description is presented herein for purposes of illustration and explanation only and is not intended to limit the present application.
It should be noted that the following description order of the embodiments is not intended to limit the preferred order of the embodiments of the present application. In the following embodiments, the descriptions of the embodiments are focused on, and for the part that is not described in detail in a certain embodiment, reference may be made to the related descriptions of other embodiments.
As shown in fig. 1, the intelligent network vehicle obtains vehicle sensing information through a sensor, and realizes the processing of the sensing information through a vehicle sensing system controller, and realizes the auxiliary or direct control of vehicle driving based on the processing result, wherein the control modes include, but are not limited to, vehicle speed control and rotation speed control, and the sensor comprises: cameras (vehicle cameras), lidar, millimeter wave radar (radar), and inertial navigation systems (IMU). It should be noted that, fig. 1 shows a method that can be achieved in the prior art, and any prior art that can implement the functions of the vehicle sensing system controller may be used as an implementation manner.
As shown in fig. 2-3, the first aspect of the present application provides a vehicle obstacle avoidance control system that merges road monitoring, and can perform better vehicle obstacle avoidance according to special conditions of a special road section (such as surrounding schools, traffic accident high-speed road sections and covered road sections), bad weather, and other situations, and mainly includes the following modules.
The road acquisition module is at least used for acquiring road environment information, and the road environment information comprises road video information.
The road information processing module is at least used for performing perception processing on the road environment information to obtain information such as category, position, speed, direction, size and time stamp of the obstacle, and obtaining road obstacle avoidance information based on the information such as category, position, speed, direction, size and time stamp of the obstacle. Specifically, obtaining the category of the obstacle through a hidden Markov model; obtaining a position and speed estimation value through a Kalman filter matrix; and (5) correcting the sensor data time through a satellite positioning system to obtain data time stamp information.
The vehicle information acquisition module is at least used for acquiring vehicle environment information, and specifically, the vehicle information acquisition module can acquire the vehicle environment information through acquisition of the environment information around the vehicle, wherein the vehicle information acquisition module comprises a camera, a laser radar, a millimeter wave radar, an inertial navigation module and a satellite positioning system sensing module.
And the vehicle information processing module is at least used for identifying a vehicle blind area and vehicle obstacle avoidance information based on the perception fusion of the vehicle environment information, and assisting and/or controlling the vehicle to avoid the obstacle in a path planning and control mode by fusing the vehicle obstacle avoidance information and road obstacle avoidance information.
Specifically, the identified vehicle blind zone may correspondingly increase the perceived range and perceived capability of the vehicle. In the application scene, the periphery of a vehicle body can be covered by one main laser radar and 4 blind supplementing laser radars, no dead angle full coverage of a sensing area is realized, the minimum detection distance is reduced to be within 5cm, but the 5 laser radars are high in installation cost and low in cost performance, and obstacles in the dead zone are identified through sensing fusion of a road and a vehicle.
Here, the intelligent network vehicle and road monitoring combined sensing capability can be established, the sensing range of the vehicle can be effectively enlarged, and the effect of active control or auxiliary control on the vehicle can be achieved.
As shown in fig. 2, in an embodiment of the present application, the vehicle obstacle avoidance control system further includes a cloud end, where the cloud end stores road obstacle avoidance information from the road information processing module and transmits the road obstacle avoidance information to the vehicle information processing module according to a download request condition of the vehicle information processing module, and by virtue of the advantages of low use cost, strong operation and maintenance capability and large storage space of the cloud end, the collaborative capability of two ends of a vehicle road can be improved.
In an embodiment of the application, the vehicle obstacle avoidance control system further includes an alarm module, where the alarm module is configured to send an alarm before the vehicle information processing module assists or controls the vehicle to avoid the obstacle.
Through the mode, an alarm or a prompt can be sent out before a vehicle blind area and a vehicle avoid obstacle or after the vehicle avoid obstacle information and the road avoid obstacle information are fused, and the prompt type can be voice or a vehicle-machine dialect popup window so as to effectively remind a driver or a user.
As shown in fig. 3, in an embodiment of the present application, the road information processing module includes a road information sensing module and a road information screening module, where the road information sensing module is configured to generate a first attribute of an obstacle based on the road environment information, and the road information screening module screens out road obstacle avoidance information from the road environment information based on the first attribute of the obstacle.
Specifically, the first attribute of the obstacle generated by the road information sensing module specifically includes information such as category, position, speed, timestamp, and the like of the obstacle. The obstacle first attribute may be understood as an obstacle attribute perceived by road monitoring.
Specifically, the road information screening module may be specifically a filter, and the perceived data required by the vehicle may be screened out by the filter. Specifically, the vehicle arrives at a road section that can be fused with the road perception, and the cloud road data is obtained by sending the data required by the request to the cloud, and the inputs required by the fusion perception algorithm are screened, for example: the camera image related data corresponding information comprises a time stamp, an image number, an image format and compressed image data, wherein the Topic is sensor/camera/image/compressed.
As shown in fig. 3, in an embodiment of the present application, the road information processing module further includes a road information recording module and a road information uploading module, where the road information recording module is configured to record the road obstacle avoidance information and the first attribute of the obstacle, and the road information uploading module is configured to upload the recorded road obstacle avoidance information to the cloud. The road obstacle avoidance information is sent to the road recording module by the road information screening module, and in this way, the road obstacle avoidance information can be recorded by the road recording module, so that the data loss of the road obstacle avoidance information is avoided. The recording module can record information according to the specified request, package the information into one or more packets, and has the functions of data packet compression and decompression.
As shown in fig. 3, in an embodiment of the present application, the vehicle information processing module further includes a vehicle information downloading module, a vehicle information recording module and a perception fusion module, where the vehicle information downloading module is configured to download the road obstacle avoidance information, and the vehicle information recording module is configured to record the road obstacle avoidance information transmitted from the vehicle downloading module and transmit the road obstacle avoidance information to the perception fusion module. The recording module can record information according to the specified request, package the information into one or more packets, and simultaneously decompress the data packets of other cloud nodes, process the data, compress and package the data, and transmit the data to the cloud.
As shown in fig. 3, in an embodiment of the present application, the vehicle information processing module further includes a vehicle information sensing module, a vehicle information screening module, and an obstacle avoidance result generating module;
the vehicle information sensing module generates obstacle second attributes and obstacle confidence based on the vehicle environment information, and generates vehicle blind zone obstacle information and vehicle obstacle avoidance information based on the obstacle second attributes and the obstacle confidence; the obstacle second attribute refers to the obstacle first attribute, and includes, for example, information of an obstacle category, a position, a speed, a time stamp, and the like. The obstacle second attribute may be understood as an obstacle attribute perceived by the vehicle, and the obstacle confidence may be used to determine that the target is the presence and confidence of the obstacle.
Preferably, the vehicle information screening module is located between the vehicle information recording module and the perception fusion module, and is used for screening road obstacle avoidance information recorded and transmitted by the vehicle information recording module, and transmitting the road obstacle avoidance information to the perception fusion module;
the perception fusion module is used for fusing the vehicle obstacle avoidance information and the road obstacle avoidance information to generate an obstacle avoidance result.
As shown in fig. 4, by adopting a multi-mode fusion algorithm combining multiple fusion algorithms (pre-fusion, intermediate fusion and post-fusion), road obstacle avoidance information and vehicle obstacle avoidance information are classified according to the data level, the feature level and the target level, and different fusion modes are adopted for fusion according to different confidence degrees of different levels and different vehicle communication conditions, so that the purpose of obstacle detection is achieved.
Specifically, the road obstacle avoidance information comprises a roadside data layer, a roadside feature layer and a roadside target layer.
The vehicle obstacle avoidance information comprises a vehicle edge data layer, a vehicle edge feature layer and a vehicle edge target layer.
The vehicle information processing module assists and/or controls a vehicle to avoid an obstacle based on a front fusion of the roadside data layer and the roadside data layer, based on a rear fusion of the roadside target layer and the roadside target layer, and based on an intermediate fusion comprising the roadside feature layer and comprising the roadside feature layer.
Specifically, pre-fusion is fusion of data levels, where vehicle data is fused at the data level, and camera data is fused at the data level and feature level, by spatial alignment and projection of the original data level to directly fuse the data in each modality. The vehicle data selects to use laser radar, millimeter wave radar or camera data according to the self-perception capability and the confidence of the sensor detection data;
the post fusion is the fusion of the target level, the vehicle sensing result and the road camera sensing result are fused, and the final sensing result is obtained by combining the confidence coefficient;
the intermediate fusion is divided into asymmetric fusion and depth fusion;
asymmetric fusion is the fusion of target level information from one branch with data level or feature level information from other branches. Asymmetric fusion has at least one branch dominant, provides target advice, and other branches provide auxiliary information. The depth fusion is to sense fusion data at a feature level for vehicles and fuse road camera data at a data level and the feature level;
and selecting a fusion method according to the confidence coefficient and/or the communication resource:
selection 1: based on the purpose of saving communication resources, if the confidence coefficient is higher, the selection priority of different multi-mode fusion algorithms is post fusion > pre fusion > asymmetric fusion > deep fusion;
selection 2: based on safety priority, on the premise of self-vehicle controller resources and calculation power permission, preferentially selecting a fusion algorithm with high confidence;
the obstacle avoidance result generation module is used for avoiding the obstacle based on the obstacle avoidance result.
In this way, the effect of early warning and controlling the vehicle can be realized through the perception range of the vehicle and the confidence range of the obstacle, and especially, the fusion obstacle avoidance can be realized through specific scenes with requirements, so that the safety of the vehicle and the personnel is ensured.
As shown in fig. 5, a second aspect of the present application provides a vehicle obstacle avoidance control method for fusing road monitoring, including: step S11: collecting road environment information, wherein the road environment information comprises road video information;
step S12: performing sensing processing on the road environment information to obtain road obstacle avoidance information; specifically, the information of the category, the position, the speed, the direction, the size, the time stamp and the like of the target can be obtained by sensing the road environment information, and the road obstacle avoidance information is obtained based on the information of the category, the position, the speed, the direction, the size, the time stamp and the like of the target.
Step S13: acquiring vehicle environment information; specifically, the vehicle information acquisition module can acquire vehicle environment information through acquisition of environment information around the vehicle, wherein the vehicle information acquisition module comprises a camera, a laser radar, a millimeter wave radar and an inertial navigation system.
Step S14: based on the sensing fusion of the vehicle environment information, the vehicle blind area and the vehicle obstacle avoidance information are identified, and the vehicle obstacle avoidance information and the road obstacle avoidance information are fused to assist and/or control the vehicle to avoid the obstacle in a path planning and control mode.
Further, the cloud end stores road obstacle avoidance information, transmits the road obstacle avoidance information to the vehicle end based on the downloading request condition of the vehicle end, recognizes a vehicle blind area and vehicle obstacle avoidance information based on sensing fusion of the vehicle environment information at the vehicle end, and assists and/or controls the vehicle to avoid the obstacle in a path planning and control mode by fusing the vehicle obstacle avoidance information and the road obstacle avoidance information.
Here, the intelligent network vehicle and road monitoring combined sensing capability can be established, the sensing range of the vehicle can be effectively enlarged, and the effect of active control or auxiliary control on the vehicle can be achieved.
As shown in fig. 6, a third aspect of the present application provides an evaluation method based on the vehicle obstacle avoidance control system or control method of the above-described embodiment of any one of the embodiments, and development preparation of the evaluation method includes configuring an intelligent network vehicle having a camera, a laser radar, a millimeter wave radar, an inertial navigation system, and the like as shown in fig. 1, and identifying surrounding pedestrians and environments in a vehicle coverage area based on a sensor provided by the intelligent network vehicle, and the road section has clear monitoring and access rights, a wireless network with good signals, and a base system such as a cloud system, which can share road videos with the vehicle.
With continued reference to fig. 6, the evaluation method of the vehicle obstacle avoidance control system or control method with road monitoring according to any one of the embodiments includes the following steps:
step S21: and calculating first perception performance based on the obstacle avoidance information of the vehicle, wherein the first perception performance comprises accuracy rate, recall rate, accuracy rate, obstacle recognition distance, minimum classification distance and the like.
Step S22: and calculating a second perception performance based on the vehicle obstacle avoidance information and the road obstacle avoidance information, wherein the second perception performance comprises an accuracy rate, a recall rate, an accuracy rate, an obstacle recognition distance, a minimum classification distance and the like.
Step S23: comparing the first perceptual performance with the second perceptual performance to obtain an evaluation result.
The sensor detects the obstacle, and can be divided into four cases:
TP samples (with obstacles, detected by sensors);
FN sample (with obstacle, sensor does not detect obstacle);
FP samples (no obstacle, sensor detects obstacle);
TN sample (no obstacle, no obstacle detected by the sensor).
Precision (Precision): for the samples of the detected obstacle, all samples of the obstacle are detected, and the proportion of the correct samples is detected, namely the accuracy rate is that positive class is predicted as positive class/predicted positive class = TP/(tp+fp);
recall (Recall) Recall: for all real obstacle samples, the proportion of the samples with detected obstacles, namely the recall rate is that positive class is predicted as positive class/original positive class=tp/(tp+fn);
accuracy (correct rate) Accuracy: for all samples, the correct sample ratio is detected, and the accuracy is all predicted correct samples/total samples= (tp+tn)/(tp+fn+fp+tn).
As shown in fig. 7, the present invention uses the average accuracy index (AP Average Precision) to evaluate perceived performance, AP IoU＝ 0.7 (Average Precision at IoU =0.7), which indicates that the average accuracy at different Recall is measured at a fixed IoU (Intersection-over-Union) threshold of 0.7. Multiple sets of precision data points are obtained by traversing different detection score thresholds, and the data points construct a curve, and the area under the curve is defined as AP IoU =0.7. The high recovery still can keep very high precision, and the larger the area under the curve is, the better the perception performance is. Under the same working condition, the accuracy rate and recall rate of the test sample of RVF and VF are counted, an accuracy rate-recall rate curve is drawn, the larger the area is compared with the area of the curve of RVF and VF and the abscissa, the better the perception performance is, wherein the threshold value is 0.7, which is only one range under one implementation mode, can be obtained through a plurality of measurement modes, and the application is not limited to the method.
In an embodiment of the present application, as shown in fig. 8-10, the evaluation method may perform a test in a severe weather scenario; the sensor quantization level of the severe weather scene is set, so that the influence degree of the severe weather on the sensor can be measured.
Specifically, the sensor quantization levels are shown in table 1:
TABLE 1
Specifically, the effect of different weather on different sensors is shown in table 2:
TABLE 2
In bad weather as described above, the same road section is subjected to obstacle type (lane line, traffic light, vehicle type, pedestrian, bicycle, small animal, etc.) identification test, and the average accuracy index AP is compared IoU ＝0.7。
And respectively carrying out a barrier position identifying test on the same road section under severe weather as described above, and comparing the first perception performance with the second perception performance to obtain a verification result.
And respectively carrying out obstacle recognition speed test on the same road section under severe weather as described above, and comparing the first perception performance with the second perception performance to obtain a verification result.
Further, for road tests with vehicle sensing blind areas, namely vehicle blind areas, green plants or buildings shield the vision of drivers or influence the working conditions of the vehicle sensing range, pedestrians pass through intention, and the road monitoring is fused with the comparison of the sensing performance sensed by the vehicles. Determining the position of an obstacle at the edge of a dead zone of a vehicle, fusing vehicle perception data with road monitoring data to obtain information such as category, position, speed and the like of the obstacle, and comparing the first perception performance with the second perception performance to obtain a verification result. Fig. 8 is a graph of road test performance results for the first perceived performance a, fig. 9 is a graph of road test performance results for the second perceived performance B, and fig. 10 is a graph of performance comparisons for the first perceived performance a and the second perceived performance B. As can be seen from fig. 10, the sensing range of the second sensing performance B is wider, it can sense the obstacle vehicle in the dead zone of the vehicle and can be obtained through road monitoring, at least this mode can be used to avoid the condition that the obstacle vehicle is not recognized due to the shielding of the tree, in an application scenario, if the obstacle vehicle is out of control to cause the brake failure, when impacting the direction of the vehicle at a high speed, other modes such as selecting high speed avoidance can be avoided to realize emergency avoidance in this mode, and by virtue of the function of automatic obstacle avoidance existing in the prior art, the driving safety of the user can be effectively ensured.
Specifically, the evaluation method can compare the difference between the vehicle perception and the vehicle perception under the condition of only vehicle perception and fusion road monitoring so as to ensure the safety of pedestrians and driving safety in special road sections and special weather.
A fourth aspect of the present application provides an evaluation system of a vehicle obstacle avoidance control system based on the fusion road monitoring according to the first aspect of the present application, including:
the first perception calculation module is at least used for calculating first perception performance based on the vehicle obstacle avoidance information;
the second perception calculation module is at least used for calculating second perception performance based on the vehicle obstacle avoidance information and the road obstacle avoidance information; and the perception comparison module is at least used for comparing the first perception performance with the second perception performance to obtain a verification result.
A fifth aspect of the present application provides a readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method of any of the above embodiments. The computer readable storage medium may include: any entity or device capable of carrying a computer program, a recording medium, a USB flash disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a software distribution medium, and so forth. The computer program comprises computer program code. The computer program code may be in the form of source code, object code, executable files, or in some intermediate form, among others. The computer readable storage medium may include: any entity or device capable of carrying computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a software distribution medium, and so forth.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, system that includes a processing module, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (12)
1. The utility model provides a vehicle obstacle avoidance control system of fusion road control which characterized in that includes:
the road information acquisition module is at least used for acquiring road environment information;
the road information processing module is at least used for processing the road environment information to obtain road obstacle avoidance information, wherein the road obstacle avoidance information comprises a roadside data layer, a roadside characteristic layer and a roadside target layer;
the vehicle information acquisition module is at least used for acquiring vehicle environment information;
the vehicle information processing module is at least used for generating vehicle obstacle avoidance information based on the vehicle environment information, fusing the vehicle obstacle avoidance information and the road obstacle avoidance information to assist and/or control the vehicle so as to avoid the obstacle, and the vehicle obstacle avoidance information comprises a vehicle side data layer, a vehicle side characteristic layer and a vehicle side target layer.
2. The vehicle obstacle avoidance control system incorporating road monitoring of claim 1 wherein the fusing of the vehicle obstacle avoidance information and the road obstacle avoidance information assists and/or controls a vehicle to avoid an obstacle, comprising:
the vehicle information processing module assists and/or controls a vehicle to avoid an obstacle based on a front fusion of the roadside data layer and the roadside data layer, based on a rear fusion of the roadside target layer and the roadside target layer, and based on an intermediate fusion including the roadside feature layer or the roadside feature layer.
3. The vehicle obstacle avoidance control system of claim 1, further comprising a cloud end for storing the obstacle avoidance information.
4. The vehicle obstacle avoidance control system of claim 1 wherein the road information processing module comprises a road information perception module and a road information screening module, the road information perception module configured to generate an obstacle first attribute based on the road environment information, and screen the road obstacle avoidance information from the road environment information based on the obstacle first attribute via the road information screening module.
5. The vehicle obstacle avoidance control system of claim 4, wherein the road information processing module further comprises a road information recording module and a road information uploading module, wherein the road information recording module is used for recording the road obstacle avoidance information, and the road information uploading module is used for uploading the recorded road obstacle avoidance information to the cloud.
6. The vehicle obstacle avoidance control system of claim 1, wherein the vehicle information processing module comprises a vehicle information downloading module, a vehicle information recording module and a perception fusion module, wherein the vehicle information downloading module is used for downloading the road obstacle avoidance information, the vehicle information recording module is used for recording the road obstacle avoidance information, and the perception fusion module is used for fusing the vehicle obstacle avoidance information and the road obstacle avoidance information to generate an obstacle avoidance result.
7. The vehicle obstacle avoidance control system of fusion road monitoring according to claim 6, wherein the vehicle information processing module further comprises a vehicle information sensing module, a vehicle information screening module and an obstacle avoidance result generating module;
the vehicle information sensing module generates a second attribute of an obstacle and an obstacle confidence degree based on the vehicle environment information, and generates a vehicle blind area and vehicle obstacle avoidance information based on the second attribute of the obstacle and the obstacle confidence degree;
the vehicle information screening module is used for recording and transmitting road obstacle avoidance information from the vehicle information recording module and transmitting the road obstacle avoidance information to the perception fusion module;
the obstacle avoidance result generation module is used for avoiding the obstacle based on the obstacle avoidance result.
8. The vehicle obstacle avoidance control method integrating road monitoring is characterized by comprising the following steps of:
collecting road environment information;
processing the road environment information to obtain road obstacle avoidance information;
acquiring vehicle environment information;
and generating vehicle obstacle avoidance information based on the vehicle environment information, and fusing the vehicle obstacle avoidance information and the road obstacle avoidance information to control the vehicle to avoid the obstacle.
9. A method of evaluating a vehicle obstacle avoidance control system based on fused roadway monitoring as claimed in any one of claims 1 to 7, comprising:
calculating a first perceived performance based on the vehicle obstacle avoidance information;
calculating a second perceived performance based on the vehicle obstacle avoidance information and the road obstacle avoidance information;
comparing the first perceived performance and the second perceived performance to obtain a verification result.
10. The evaluation method according to claim 9, characterized by comprising:
the evaluation method is used for testing in severe weather scenes;
setting a sensor quantization level of a bad weather scene, wherein the sensor quantization level is used for measuring the influence degree of bad weather on a sensor;
and performing obstacle position and/or speed recognition tests on the same road section based on the sensor quantization level so as to compare the first perception performance with the second perception performance to obtain a verification result.
11. An evaluation system of a vehicle obstacle avoidance control system based on fused road monitoring as claimed in any one of claims 1 to 7, comprising:
the first perception calculation module is at least used for calculating first perception performance based on the vehicle obstacle avoidance information;
the second perception calculation module is at least used for calculating second perception performance based on the vehicle obstacle avoidance information and the road obstacle avoidance information;
and the perception comparison module is at least used for comparing the first perception performance with the second perception performance to obtain a verification result.
12. A readable storage medium storing a computer program, which when executed by a processor performs the steps of the method according to any one of claims 8-10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311704926.4A CN117681865A (en) | 2023-12-12 | 2023-12-12 | Vehicle obstacle avoidance control system integrating road monitoring and evaluation method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311704926.4A CN117681865A (en) | 2023-12-12 | 2023-12-12 | Vehicle obstacle avoidance control system integrating road monitoring and evaluation method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117681865A true CN117681865A (en) | 2024-03-12 |
Family
ID=90129704
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311704926.4A Pending CN117681865A (en) | 2023-12-12 | 2023-12-12 | Vehicle obstacle avoidance control system integrating road monitoring and evaluation method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117681865A (en) |
-
2023
- 2023-12-12 CN CN202311704926.4A patent/CN117681865A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220157092A1 (en) | Black box data management method, apparatus, and device for intelligent driving vehicle | |
| CN113284366B (en) | Vehicle blind area early warning method, early warning device, MEC platform and storage medium | |
| US20190265712A1 (en) | Method for determining driving policy | |
| CN109345829B (en) | Unmanned vehicle monitoring method, device, equipment and storage medium | |
| CN113240909B (en) | Vehicle monitoring method, equipment, cloud control platform and vehicle road cooperative system | |
| KR101103526B1 (en) | Collision Avoidance Method Using Stereo Camera | |
| US20210303883A1 (en) | Deterioration diagnosis device, deterioration diagnosis system, deterioration diagnosis method, and storage medium for storing program | |
| JP6418574B2 (en) | Risk estimation device, risk estimation method, and computer program for risk estimation | |
| WO2021155685A1 (en) | Map updating method, apparatus and device | |
| JP6962604B2 (en) | Collaborative blindspot alerting methods and equipment for inter-vehicle communication infrastructure with fault tolerance and fracture robustness in extreme situations | |
| CN111582130B (en) | Traffic behavior perception fusion system and method based on multi-source heterogeneous information | |
| CA3056611A1 (en) | Automatic warning generation system intended for the users of a road | |
| US11335136B2 (en) | Method for ascertaining illegal driving behavior by a vehicle | |
| CN108958248A (en) | Standby system | |
| CN112598908B (en) | Driver red light running recognition method and device, electronic equipment and storage medium | |
| CN111277956A (en) | Method and device for collecting vehicle blind area information | |
| KR101731789B1 (en) | ADAS controlling method using road recognition and control system | |
| CN117681865A (en) | Vehicle obstacle avoidance control system integrating road monitoring and evaluation method thereof | |
| CN116630891A (en) | Traffic abnormal event detection system and method | |
| Shanshan et al. | An evaluation system based on user big data management and artificial intelligence for automatic vehicles | |
| CN115240470A (en) | NR-V2X-based weak traffic participant collision early warning system and method | |
| KR102121423B1 (en) | Server and method for recognizing road line using camera image | |
| CN113954826B (en) | Vehicle control method and system for vehicle blind area and vehicle | |
| WO2022196660A1 (en) | Driving assistance device, driving assistance method, drive recorder, and driving assistance control program | |
| WO2023087182A1 (en) | Information interaction method and device and system |
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 |