CN111985373A - Safety early warning method and device based on traffic intersection identification and electronic equipment - Google Patents
Safety early warning method and device based on traffic intersection identification and electronic equipment Download PDFInfo
- Publication number
- CN111985373A CN111985373A CN202010808246.7A CN202010808246A CN111985373A CN 111985373 A CN111985373 A CN 111985373A CN 202010808246 A CN202010808246 A CN 202010808246A CN 111985373 A CN111985373 A CN 111985373A
- Authority
- CN
- China
- Prior art keywords
- traffic
- intersection
- identification
- recognition
- crossing
- 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
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000012549 training Methods 0.000 claims abstract description 43
- 238000005286 illumination Methods 0.000 claims abstract description 16
- 238000010801 machine learning Methods 0.000 claims abstract description 14
- 230000008569 process Effects 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 238000000547 structure data Methods 0.000 claims description 13
- 238000009401 outcrossing Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 8
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 2
- 239000002131 composite material Substances 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/582—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of traffic signs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/24—Reminder alarms, e.g. anti-loss alarms
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
- G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Atmospheric Sciences (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a safety early warning method, a safety early warning device and electronic equipment based on traffic intersection identification, wherein the method comprises the following steps: acquiring road surface video data in front of a vehicle in real time; converting road surface video data into a plurality of pictures to be identified; inputting a plurality of pictures to be recognized into a preset traffic sign recognition model, obtaining a recognition result of an intersection sign, and determining whether a vehicle is about to arrive at or drive away from the traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection sign exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks; if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information; the traffic sign recognition model is obtained by taking the image characteristics and the intersection sign characteristics determined according to the road surface picture samples under different weather or illumination conditions as input and machine learning training.
Description
Technical Field
The invention relates to the technical field of road traffic safety, in particular to a safety early warning method and device based on traffic intersection identification and electronic equipment.
Background
The traffic intersection is an important scene in the whole process of road traffic, and because the road conditions of the traffic intersection are complex and changeable, the collected traffic flow is numerous, and the distribution of vehicles and pedestrians is denser, more traffic conflicts and potential safety hazards exist at the traffic intersection.
For the safety management of a traffic intersection, in the prior art, two modes are generally adopted, wherein one mode is that a monitoring and reminding system is installed at the traffic intersection by means of peripheral equipment, and when the monitoring and reminding system monitors that a vehicle reaches the intersection, an acousto-optic signal is sent out to remind a driver in the vehicle, and in the mode, the driver in the vehicle cannot receive the acousto-optic reminding signal due to interference caused by dense people and vehicles at the intersection or noisy environment; secondly, a monitoring reminding system is installed on a vehicle, the characteristic pictures of all intersections are adopted to identify the traffic intersections, and then the drivers are reminded through sound and light signals.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a safety early warning method and device based on traffic intersection identification and electronic equipment.
In a first aspect, an embodiment of the present invention provides a traffic intersection identification-based safety early warning method, including:
acquiring road surface video data in front of a vehicle along a driving route in real time;
converting the road surface video data into a plurality of pictures to be identified;
inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of an intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input.
Further, the training process of the traffic sign recognition model comprises the following steps: and (3) detecting and training the road surface pictures under different weather or illumination conditions by adopting Python language and PyTorch frame, and obtaining a model data file.
Further, the safety precaution method further comprises the following steps: and compressing and encrypting the model data file to obtain a network structure data file and a weight parameter data file, wherein the network structure data file is a network structure storage file of the road surface picture data required by the model training process, and the weight parameter data file is weight parameter information data which is obtained after the model training and is used for describing the model.
Further, the network structure data file adopts a net file format, and the weight parameter data file adopts a weight file format.
Further, the safety early warning method converts the road surface video data into a plurality of pictures to be identified through an executable program and a dynamic link library.
Further, the model data file also comprises category information, position information and confidence information of the intersection identification;
correspondingly, when the identification result is that the intersection identification exists, the identification result further comprises category information, position information and confidence degree information corresponding to the intersection identification.
Further, the safety early warning information is information including an identification result or early warning prompt information corresponding to the identification result.
In a second aspect, an embodiment of the present invention further provides a safety precaution device based on traffic intersection identification, including:
the video acquisition module is used for acquiring road surface video data in front of the vehicle along the driving route in real time;
the data processing module is used for converting the road surface video data into a plurality of pictures to be identified;
the recognition judging module is used for inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of the intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
the early warning generation module is used for generating safety early warning information if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, and the pattern features and the crossing sign features determined according to road surface picture samples under different weather or illumination conditions are used as input.
In a third aspect, an embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor executes the computer program to implement the above-mentioned safety precaution method based on traffic intersection identification.
In a fourth aspect, the embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the traffic intersection identification-based safety precaution method as described above.
The embodiment of the invention provides a traffic intersection identification-based safety early warning method, a device and electronic equipment, wherein the acquired road surface video data is converted into a plurality of pictures to be identified, the pictures to be identified are input into a preset traffic identification model for identification, the accuracy of the obtained identification result is high, whether a vehicle is about to arrive at or drive away from a traffic intersection at the moment is determined according to the identification result, and finally, safety early warning information is generated to remind a driver while the vehicle is determined to arrive at or drive away from the traffic intersection, so that the driver has more sufficient time to observe road conditions, and the operations such as deceleration, parking and the like are performed in time, the driving safety is improved, and the incidence rate of driving accidents at the traffic intersection is effectively reduced.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flow chart of a safety warning method based on traffic intersection identification according to an embodiment of the present invention;
fig. 2 is a schematic view of an application scenario of a safety early warning method based on traffic intersection identification according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a safety precaution device based on traffic intersection identification according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Reference numerals:
210: a vehicle-mounted intelligent terminal; 220: an external camera;
310: a video acquisition module; 320: a data processing module; 330: a recognition and judgment module; 340: an early warning generation module;
410: a processor; 420: a communication interface; 430: a memory; 440: a communication bus.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention provides a traffic intersection identification-based safety early warning method, a traffic intersection identification-based safety early warning device and electronic equipment, wherein the traffic intersection identification-based safety early warning method specifically comprises the following steps:
acquiring road surface video data in front of a vehicle along a driving route in real time;
converting the road surface video data into a plurality of pictures to be identified;
inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of an intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input.
According to the traffic intersection identification-based safety early warning method provided by the embodiment of the invention, the acquired road surface video data is converted into a plurality of pictures to be identified, the pictures to be identified are input into the preset traffic identification model for identification, the accuracy of the obtained identification result is high, whether the vehicle is about to arrive at or drive away from the traffic intersection at the moment is determined according to the identification result, and finally safety early warning information is generated to remind a driver while the vehicle is about to arrive at or drive away from the traffic intersection, so that the driver has more sufficient time to observe road conditions, and the operations such as speed reduction, parking and the like are carried out in time, so that the driving safety is improved, and the occurrence rate of driving accidents at the traffic intersection is effectively reduced.
The safety early warning method based on traffic intersection identification provided by the embodiment of the invention is explained in more detail below.
In a first aspect, an embodiment of the present invention provides a traffic intersection identification-based security early warning method, fig. 1 is a schematic flow diagram of a traffic intersection identification-based security early warning method provided in an embodiment of the present invention, and fig. 2 is a schematic application scenario diagram of a traffic intersection identification-based security early warning method provided in an embodiment of the present invention, where the method may be applied to an intelligent terminal device that implements a traffic intersection identification function, such as a vehicle-mounted intelligent terminal, as shown in fig. 1 to fig. 2, and the method includes:
step S110, road surface video data in front of the vehicle along the driving route is acquired in real time.
When the method is applied to the vehicle-mounted intelligent terminal for realizing the traffic intersection identification function, the road surface video data can be directly obtained in real time through the vehicle-mounted intelligent terminal with the camera shooting unit inside, and at the moment, the vehicle-mounted intelligent terminal needs to be hung above a front windshield in a vehicle, such as one side of a vehicle traveling data recorder. The vehicle-mounted intelligent terminal embedded with the method can also be connected with an external camera or a driving recorder, the external camera or the driving recorder assists the vehicle-mounted intelligent terminal to acquire road surface video data in front of the vehicle, and at the moment, the vehicle-mounted intelligent terminal can be mounted at any suitable position in the vehicle and can be specifically designed according to actual application scenes.
In this embodiment, fig. 2 only shows an application scenario in which the vehicle-mounted intelligent terminal and the external camera work together, and with reference to fig. 1 and fig. 2, the external camera 220 dynamically monitors a road surface state in front of the vehicle along a driving route, and acquires current road surface video data in front of the vehicle in real time to transmit to the vehicle-mounted intelligent terminal 210 inside the vehicle.
And step S120, converting the road surface video data into a plurality of pictures to be identified.
The vehicle-mounted intelligent terminal 210 converts the current road surface video data in front of the vehicle, which is obtained through one or more external cameras, into a plurality of pictures to be recognized in real time.
Step S130, further including step S1301 and step S1302, wherein step S1301 is to input the multiple pictures to be recognized into a preset traffic identification recognition model, and obtain a recognition result of an intersection identifier output by the traffic identification recognition model; the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input; step S1302, determining whether the vehicle is about to arrive at or leave the traffic intersection according to the identification result; wherein the identification result comprises whether the intersection identification exists or not; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
the vehicle-mounted intelligent terminal 210 recognizes the plurality of converted pictures to be recognized through a preset traffic identification recognition model, and accurately compares each picture to be recognized with a picture with or without an intersection identification in the traffic identification recognition model to obtain a recognition result of whether the intersection identification exists.
The crossing mark comprises a crossing pattern mark, a figure mark and a simple marking mark, and specifically comprises a pedestrian crossing mark (pedestrian crossing white line), a pedestrian crossing forenotice mark (pedestrian crossing forenotice figure), a crossing deceleration line mark (deceleration strip), a guide arrow mark (left-right turning, turning white line and guide icon) and any marks with traffic crossing identification and guide functions. Correspondingly, the identification result of each picture to be identified specifically includes: the existence of pedestrian crossing identification indicates that pedestrian paths exist; the absence of pedestrian crossing identification indicates that no pedestrian path exists; the pedestrian crosswalk forenotice mark indicates that the pedestrian crosswalk is about to be encountered; the pedestrian crosswalk forenotice mark does not exist, and the pedestrian crosswalk is not met; the existence of the crossing deceleration line mark indicates that the traffic crossing is encountered; the crossing deceleration line mark does not exist, which indicates that the traffic crossing is not met; the presence of the guide arrow mark indicates that the vehicle is about to meet the intersection or just drive away from the intersection (corresponding to the state of just driving away from the intersection when the turning or turning motion is finished), and the absence of the guide arrow mark indicates that the vehicle is not meeting the intersection or just does not drive away from the intersection. Of course, the intersection identifier corresponding to the state where the vehicle has reached the intersection is the same as or similar to the above-mentioned various intersection identifiers, and the identification principle is the same, which is not described herein again.
In the construction process of the traffic sign recognition model in the early preset stage, a large number of road surface pictures with or without crossing signs need to be collected as samples, the image characteristics and crossing sign characteristics determined according to the large number of road surface picture samples are used as input, the traffic sign recognition model is constructed through machine learning training by utilizing a deep learning algorithm, and the specific training process is as follows:
1-1, acquiring a training data set for model training, wherein the training data set comprises a large number of road surface pictures under different weather or illumination conditions, and the road surface pictures can select a large number of road surface pictures with or partially with pedestrian crossing marks/pedestrian crossing forenotice marks/intersection deceleration line marks/guide arrow marks; a large number of road surface pictures without crosswalk marks/crosswalk forenotice marks/intersection deceleration line marks/guide arrow marks can be selected, so that a training data set is enriched; after a large number of road surface pictures in the training data set for model training are obtained, picture labeling and intersection identification category division are carried out on each road surface picture by combining with actual conditions, and in the process, data picture optimization processing operations such as useless picture data elimination and picture data enhancement are carried out to improve the accuracy and usability of the whole training data set.
1-2, taking a large number of road surface pictures of the training data set obtained in the previous step as samples, taking image characteristics and intersection identification characteristics determined according to the large number of road surface picture samples as input data used for model training, training by adopting a deep learning algorithm, and performing targeted training by using a specified language and a specified frame, thereby constructing a traffic identification recognition model for recognizing that a vehicle is about to arrive at an intersection.
Step S140, if the judgment result of the step S1302 is that the vehicle is about to arrive at or drive away from the traffic intersection, generating safety early warning information; otherwise, the process returns to step S110.
That is, the vehicle-mounted intelligent terminal 210 continues to execute step S140 according to the determination result of step S1302, and if any one or more of the crosswalk identifier, the crosswalk forenotice identifier, the intersection deceleration line identifier and the guide arrow identifier exists in the to-be-recognized picture, that is, when it is determined that the vehicle is about to arrive at or drive away from the traffic intersection, generates safety warning information to remind the driver, where the safety warning information is information including the recognition result, or warning prompt information corresponding to the recognition result, such as text message prompt, voice broadcast prompt, flash warning prompt and the like.
According to the method, the vehicle-mounted intelligent terminal 210 converts road surface video data dynamically acquired by the external camera 220 in real time into a plurality of pictures to be recognized, then the pictures to be recognized are input into a preset traffic sign recognition model for recognition, whether a traffic sign or a marking indicating that the vehicle is about to arrive/has arrived or just drives away from a traffic intersection exists on the current road surface is recognized and judged, then whether the vehicle is about to arrive or drives away from the traffic intersection is determined according to a recognition result, finally, safety early warning information is generated while the vehicle is determined to arrive or drive away from the traffic intersection, a driver is reminded in various modes through the intelligent vehicle-mounted intelligent terminal, the driver has more sufficient time to observe road conditions, operations such as speed reduction and parking are carried out in time, driving safety is improved, and the occurrence rate of driving accidents at the traffic intersection is effectively reduced.
On the basis of the above embodiment, the training process of the traffic sign recognition model includes: and (3) detecting and training the road surface pictures under different weather or illumination conditions by adopting Python language and PyTorch frame, and obtaining a model data file. The model data file is actually a composite data file in which the network structure data and the weight parameter data are mixed. The method comprises the steps of performing machine learning training by using a deep learning algorithm, optimizing and recreating a mainstream model, reducing the size of the model as much as possible under the condition of ensuring reasoning precision on an embedded platform, appointing to use Python language and apply a PyTorch frame to perform detection training on a large number of road pictures under different weather or illumination conditions, obtaining a comprehensive model data file, continuously enriching a training data set, and continuing training, thereby increasing the robustness of the model. The Python language is a cross-platform computer programming language and is also an object-oriented dynamic type language, and the PyTorch framework is a machine learning library of an open-source Python language.
On the basis of the above embodiment, the safety precaution method further includes: and compressing and encrypting the model data file to obtain a network structure data file and a weight parameter data file, wherein the network structure data file is a network structure storage file of road surface picture data required in the model training process, and the weight parameter data file is weight parameter information data which is obtained after the model training and is used for describing the model and can also be understood as bias data corresponding to the model. The network structure data file adopts a net file format, and the weight parameter data file adopts a weight file format. The network structure data file (net file) and the weight parameter data file (weight file) are both binary self-defined private files and are proprietary formats for realizing the traffic intersection identification-based safety early warning method through a traffic identification model. And the network structure data file (. net file) is a file saved by a network for training, and the weight parameter data file (. weight) is a weight parameter data file obtained by training by using a training network, and both the weight parameter data file and the weight parameter data file are used in the identification process of whether the vehicle is about to arrive at or drive away from the intersection, so that the hardware requirement in the identification process is reduced, and the hardware resource is saved.
On the basis of the above embodiment, the safety early warning method converts the road surface video data into a plurality of pictures to be recognized through an executable program and a dynamic link library. Specifically, an executable program and a dynamic link library which are available for the embedded platform are generated by using a compiling or cross-compiling mode in an inference process of the identification and judgment of the traffic identification model, and a road video acquired by an external camera in real time is converted into a plurality of pictures to be identified through the executable program and the dynamic link library.
On the basis of the above embodiment, the model data file further includes category information, position information, and confidence information of the intersection identifier; correspondingly, when the identification result is that the intersection identification exists, the identification result further comprises category information, position information and confidence degree information corresponding to the intersection identification. Specifically, the category information, the position information and the confidence information of the intersection identification are stored in a weight parameter data file (. weight file) after the encryption and compression processing of a model data file, so as to assist in describing the traffic identification recognition model. The position information refers to the accurate position of the intersection identifier in the current picture to be identified, and is embodied as x-axis coordinates and y-axis coordinates of the upper left corner of the intersection identifier, and specific parameters such as the height and the width of the current picture to be identified.
The category information is to judge the object in the position information, and the detected object is a specific intersection identification category; the confidence level information, i.e., the confidence level of the type determination in the position information, is calculated by a model.
The method obtains category information, position information and confidence information corresponding to the intersection identification in the picture to be identified by calling the network structure data file (. net file) and the weight parameter data file (. weight file) after the model data file is compressed and decrypted, specifically, the position information can call the model data file through an executable program and process the picture to be identified to obtain the accurate position of the intersection identification in the picture to be identified, and the judgment of the category information is that the intersection identification with the determined accurate position in the picture to be identified is taken as an object to judge the category of the intersection identification to which the intersection identification belongs, and the confidence information is that the confidence of the object with the determined accurate position and the determined category of the intersection identification is judged through the specific operation of the model. Correspondingly, when the identification result is that the intersection identification exists, the identification result further includes category information, position information and confidence information corresponding to the identified intersection identification, and even may further include identification time information. And then determining the intersection identification existing in the picture to be recognized according to the category information, the position information, the confidence information, the recognition time information and the like of the intersection identification corresponding to the picture to be recognized, and determining the recognition result of whether the vehicle corresponding to the picture to be recognized is about to arrive at or drive away from the intersection.
On the basis of the above embodiment, the safety warning information is information including the recognition result or warning prompt information corresponding to the recognition result, and specifically may be any prompt information capable of transmitting a safety warning signal, such as text message prompt, voice broadcast prompt, flash warning prompt, and the like, and may be sent out through a speaker or a display screen of the vehicle-mounted intelligent terminal to remind the driver of safety. Meanwhile, if the vehicle-mounted intelligent terminal executing the traffic intersection identification-based safety early warning method is connected with the remote server, the safety early warning information can be reported to the remote server so as to provide a data basis for realizing the traffic intersection identification-based safety early warning remote cooperation service.
The traffic intersection identification-based safety early warning device provided by the embodiment of the invention is described below, and the traffic intersection identification-based safety early warning device described below and the traffic intersection identification-based safety early warning method described above can be referred to correspondingly, and are not described herein again.
In a second aspect, an embodiment of the present invention further provides a safety precaution device based on traffic intersection identification, as shown in fig. 3, the device includes:
the video acquisition module 310 is used for acquiring road surface video data in front of the vehicle along the driving route in real time;
the data processing module 320 is used for converting the road surface video data into a plurality of pictures to be identified;
the recognition judging module 330 is configured to input the multiple pictures to be recognized into a preset traffic identifier recognition model, obtain a recognition result of an intersection identifier output by the traffic identifier recognition model, and determine whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, where the recognition result includes whether the intersection identifier exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
the early warning generation module 340 is configured to generate safety early warning information if it is determined that a vehicle is about to arrive at or drive away from a traffic intersection;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, and the pattern features and the crossing sign features determined according to road surface picture samples under different weather or illumination conditions are used as input.
In addition, the video capture module 310 may be a camera unit disposed inside the vehicle-mounted intelligent terminal, or may be an external camera connected to the vehicle-mounted intelligent terminal. And the data processing module 320 and the identification judging module 330 may be combined into one identification module, and the identification module can perform two corresponding functions.
In the traffic intersection recognition-based safety early warning device provided by the embodiment of the invention, the modules cooperate with each other, the data processing module 320 converts the road surface video data acquired by the video acquisition module 310 into a plurality of pictures to be recognized, and then the recognition judgment module 330 inputs the pictures to be recognized into the preset traffic identification recognition model based on deep learning for recognition, so that the accuracy of the acquired recognition result is high, and determines whether the vehicle is about to arrive at or drive away from the traffic intersection at the moment according to the recognition result, the early warning generation module 340 generates safety early warning information when the recognition judgment module 330 determines that the vehicle is about to arrive at or drive away from the traffic intersection at the moment, the system can remind the driver, so that the driver has more sufficient time to observe the road condition, and timely carry out operations such as speed reduction and parking, thereby improving the driving safety and effectively reducing the incidence rate of driving accidents at traffic intersections.
In a third aspect, an embodiment of the present invention further provides an electronic device, and fig. 4 illustrates an entity structural schematic diagram of the electronic device, and as shown in fig. 4, the electronic device may include: a processor (processor)410, a communication Interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communication Interface 420 and the memory 430 communicate with each other via the communication bus 440. The processor 410 may call logic instructions in the memory 430 to execute the safety precaution method based on traffic intersection identification provided by the above-mentioned embodiments of the method, the method comprising:
acquiring road surface video data in front of a vehicle along a driving route in real time;
converting the road surface video data into a plurality of pictures to be identified;
inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of an intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input.
In addition, the logic instructions in the memory 430 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In addition, an embodiment of the present invention further provides a computer program product, where the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, the computer can execute the safety precaution method based on traffic intersection identification provided by the above-mentioned method embodiments, and the method includes:
acquiring road surface video data in front of a vehicle along a driving route in real time;
converting the road surface video data into a plurality of pictures to be identified;
inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of an intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input.
In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented by a processor to perform the traffic intersection identification-based safety precaution method provided in the foregoing embodiments, where the method includes:
acquiring road surface video data in front of a vehicle along a driving route in real time;
converting the road surface video data into a plurality of pictures to be identified;
inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of an intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present 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 solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A safety early warning method based on traffic intersection identification is characterized by comprising the following steps:
acquiring road surface video data in front of a vehicle along a driving route in real time;
converting the road surface video data into a plurality of pictures to be identified;
inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of an intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined, generating safety early warning information;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, wherein the traffic sign recognition model is obtained by taking image features and crossing sign features determined according to road surface picture samples under different weather or illumination conditions as input.
2. The safety precaution method based on traffic intersection recognition of claim 1, wherein the training process of the traffic sign recognition model comprises: and (3) detecting and training the road surface pictures under different weather or illumination conditions by adopting Python language and PyTorch frame, and obtaining a model data file.
3. The safety precaution method based on traffic intersection recognition of claim 2, further comprising: and compressing and encrypting the model data file to obtain a network structure data file and a weight parameter data file, wherein the network structure data file is a network structure storage file of the road surface picture data required by the model training process, and the weight parameter data file is weight parameter information data which is obtained after the model training and is used for describing the model.
4. The traffic intersection identification-based safety precaution method of claim 3, wherein said network structure data file is in a net file format and said weight parameter data file is in a weight file format.
5. The traffic intersection identification-based safety precaution method according to any one of claims 1 to 4, characterized in that the road video data is converted into a plurality of pictures to be identified by an executable program and a dynamic link library.
6. The traffic intersection recognition-based safety precaution method according to claim 2, wherein the model data file further includes category information, location information and confidence information of the intersection identification;
correspondingly, when the identification result is that the intersection identification exists, the identification result further comprises category information, position information and confidence degree information corresponding to the intersection identification.
7. The safety precaution method based on traffic intersection recognition of claim 1, wherein the safety precaution information is information including recognition results or precaution prompt information corresponding to the recognition results.
8. The utility model provides a safety precaution device based on traffic crossing discernment which characterized in that includes:
the video acquisition module is used for acquiring road surface video data in front of the vehicle along the driving route in real time;
the data processing module is used for converting the road surface video data into a plurality of pictures to be identified;
the recognition judging module is used for inputting the pictures to be recognized into a preset traffic identification recognition model, obtaining a recognition result of the intersection identification output by the traffic identification recognition model, and determining whether a vehicle is about to arrive at or drive away from a traffic intersection according to the recognition result, wherein the recognition result comprises whether the intersection identification exists; the crossing marks comprise pedestrian crossing marks, pedestrian crossing forenotice marks, crossing deceleration line marks and guide arrow marks;
the early warning generation module is used for generating safety early warning information if the fact that the vehicle is about to arrive at or drive away from the traffic intersection is determined;
the traffic sign recognition model is obtained through machine learning training and is used for carrying out crossing sign recognition, and the pattern features and the crossing sign features determined according to road surface picture samples under different weather or illumination conditions are used as input.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 7 when executing the program.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for safety precaution based on traffic intersection identification according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010808246.7A CN111985373A (en) | 2020-08-12 | 2020-08-12 | Safety early warning method and device based on traffic intersection identification and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010808246.7A CN111985373A (en) | 2020-08-12 | 2020-08-12 | Safety early warning method and device based on traffic intersection identification and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111985373A true CN111985373A (en) | 2020-11-24 |
Family
ID=73434952
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010808246.7A Pending CN111985373A (en) | 2020-08-12 | 2020-08-12 | Safety early warning method and device based on traffic intersection identification and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111985373A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112735165A (en) * | 2020-12-28 | 2021-04-30 | 深圳市金溢科技股份有限公司 | Vehicle-road cooperative road early warning method, vehicle-mounted unit and road side unit |
| CN112818956A (en) * | 2021-03-19 | 2021-05-18 | 张秀霞 | Intelligent traffic data acquisition system and acquisition method thereof |
| CN113392762A (en) * | 2021-06-15 | 2021-09-14 | 北京纵目安驰智能科技有限公司 | Intersection detection method, system, terminal and computer readable storage medium |
| CN114003672A (en) * | 2021-10-22 | 2022-02-01 | 北京百度网讯科技有限公司 | Method, device, equipment and medium for processing road dynamic event |
| WO2023273259A1 (en) * | 2021-06-30 | 2023-01-05 | 北京百度网讯科技有限公司 | Method and apparatus for generating direction recognition model, and device, medium and program product |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150161455A1 (en) * | 2012-05-25 | 2015-06-11 | Denso Corporation | Intersection recognition device and road type recognition device |
| US20170190334A1 (en) * | 2016-01-06 | 2017-07-06 | GM Global Technology Operations LLC | Prediction of driver intent at intersection |
| CN107609485A (en) * | 2017-08-16 | 2018-01-19 | 中国科学院自动化研究所 | The recognition methods of traffic sign, storage medium, processing equipment |
| CN107886034A (en) * | 2016-09-30 | 2018-04-06 | 比亚迪股份有限公司 | Driving based reminding method, device and vehicle |
| CN108416283A (en) * | 2018-02-28 | 2018-08-17 | 华南理工大学 | A kind of pavement marking recognition methods based on SSD |
| CN109389838A (en) * | 2018-11-26 | 2019-02-26 | 爱驰汽车有限公司 | Unmanned crossing paths planning method, system, equipment and storage medium |
| CN111259808A (en) * | 2020-01-17 | 2020-06-09 | 北京工业大学 | Detection and identification method of traffic identification based on improved SSD algorithm |
-
2020
- 2020-08-12 CN CN202010808246.7A patent/CN111985373A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150161455A1 (en) * | 2012-05-25 | 2015-06-11 | Denso Corporation | Intersection recognition device and road type recognition device |
| US20170190334A1 (en) * | 2016-01-06 | 2017-07-06 | GM Global Technology Operations LLC | Prediction of driver intent at intersection |
| CN107886034A (en) * | 2016-09-30 | 2018-04-06 | 比亚迪股份有限公司 | Driving based reminding method, device and vehicle |
| CN107609485A (en) * | 2017-08-16 | 2018-01-19 | 中国科学院自动化研究所 | The recognition methods of traffic sign, storage medium, processing equipment |
| CN108416283A (en) * | 2018-02-28 | 2018-08-17 | 华南理工大学 | A kind of pavement marking recognition methods based on SSD |
| CN109389838A (en) * | 2018-11-26 | 2019-02-26 | 爱驰汽车有限公司 | Unmanned crossing paths planning method, system, equipment and storage medium |
| CN111259808A (en) * | 2020-01-17 | 2020-06-09 | 北京工业大学 | Detection and identification method of traffic identification based on improved SSD algorithm |
Non-Patent Citations (4)
| Title |
|---|
| GABRIEL VILLALÓN-SEPÚLVEDA 等: "Traffic Sign Detection System for Locating Road Intersections and Roundabouts: The Chilean Case", 《SENSORS》, pages 1 - 21 * |
| 任洪梅;: "基于深度学习的路面交通标志识别", 《信息通信》, no. 04, pages 26 - 27 * |
| 尹斌军: "基于卷积神经网络的交通标志检测识别算法设计与FPGA验证", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, pages 035 - 256 * |
| 李昕蔚;丁正彦;尚岩峰;祝永新;汪辉;钟雪霞;田犁;黄尊恺;封松林;: "一种针对路口监控图像的区域分割方法", 《计算机应用与软件》, vol. 37, no. 03, pages 236 - 243 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112735165A (en) * | 2020-12-28 | 2021-04-30 | 深圳市金溢科技股份有限公司 | Vehicle-road cooperative road early warning method, vehicle-mounted unit and road side unit |
| CN112735165B (en) * | 2020-12-28 | 2022-03-08 | 深圳市金溢科技股份有限公司 | Vehicle-road cooperative road early warning method, vehicle-mounted unit and road side unit |
| CN112818956A (en) * | 2021-03-19 | 2021-05-18 | 张秀霞 | Intelligent traffic data acquisition system and acquisition method thereof |
| CN113392762A (en) * | 2021-06-15 | 2021-09-14 | 北京纵目安驰智能科技有限公司 | Intersection detection method, system, terminal and computer readable storage medium |
| CN113392762B (en) * | 2021-06-15 | 2024-04-26 | 北京纵目安驰智能科技有限公司 | Intersection detection method, system, terminal and computer readable storage medium |
| WO2023273259A1 (en) * | 2021-06-30 | 2023-01-05 | 北京百度网讯科技有限公司 | Method and apparatus for generating direction recognition model, and device, medium and program product |
| CN114003672A (en) * | 2021-10-22 | 2022-02-01 | 北京百度网讯科技有限公司 | Method, device, equipment and medium for processing road dynamic event |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111985373A (en) | Safety early warning method and device based on traffic intersection identification and electronic equipment | |
| US10913461B2 (en) | Driving behavior determining method, apparatus, and device, and storage medium | |
| CN108632530B (en) | Data processing method, device and equipment for vehicle damage assessment, client and electronic equipment | |
| CN109345829B (en) | Unmanned vehicle monitoring method, device, equipment and storage medium | |
| US20210192239A1 (en) | Method for recognizing indication information of an indicator light, electronic apparatus and storage medium | |
| CN109782364B (en) | Traffic sign board missing detection method based on machine vision | |
| CN110519727B (en) | CGW (Central gateway-based) and data processing method based on CGW | |
| CN112766069A (en) | Vehicle illegal parking detection method and device based on deep learning and electronic equipment | |
| CN111967396A (en) | Processing method, device and equipment for obstacle detection and storage medium | |
| CN113221756A (en) | Traffic sign detection method and related equipment | |
| CN114677848B (en) | Perception early warning system, method, device and computer program product | |
| CN112990162A (en) | Target detection method and device, terminal equipment and storage medium | |
| KR101874352B1 (en) | VMS, In-vehicle terminal and intelligent transport system including the same for efficient transmission of traffic information on the road | |
| CN107548466B (en) | Method and device for detecting road barrier | |
| CN113052047B (en) | Traffic event detection method, road side equipment, cloud control platform and system | |
| CN114419603A (en) | Automatic driving vehicle control method and system and automatic driving vehicle | |
| CN106803937B (en) | Double-camera video monitoring method, system and monitoring device with text log | |
| CN108932503A (en) | The recognition methods of Chinese herbaceous peony obstacle and device, storage medium, terminal under bad weather | |
| CN115489536B (en) | Driving assistance method, system, equipment and readable storage medium | |
| CN116612638A (en) | Traffic collision accident detection method, device and readable medium | |
| CN113609956B (en) | Training method, recognition device, electronic equipment and storage medium | |
| CN116798247A (en) | Accident data prompting method and related equipment | |
| CN110198439A (en) | Method and apparatus for testing the image recognition performance of ADAS camera automatically | |
| KR102377393B1 (en) | Image analysis method and system for recognition of Heavy Equipment and Gas Pipe | |
| CN114037763A (en) | Target object missing detection method and device, storage medium and electronic equipment |
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 |