CN113887297A - Safe driving monitoring method and device for forming data closed loop based on cloud - Google Patents
Safe driving monitoring method and device for forming data closed loop based on cloud Download PDFInfo
- Publication number
- CN113887297A CN113887297A CN202111016009.8A CN202111016009A CN113887297A CN 113887297 A CN113887297 A CN 113887297A CN 202111016009 A CN202111016009 A CN 202111016009A CN 113887297 A CN113887297 A CN 113887297A
- Authority
- CN
- China
- Prior art keywords
- data
- behavior detection
- behavior
- abnormal behavior
- driving
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention provides a safe driving monitoring method and device for forming a data closed loop based on a cloud end, wherein the method comprises the following steps: receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; the behavior detection model is obtained based on sample training; driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data; and inputting the abnormal behavior recognition data to the equipment terminal so as to perform supplementary training on the behavior detection model of the equipment terminal. The invention carries out behavior recognition on the received abnormal behavior detection data detected by the equipment terminal based on the behavior detection model, returns the abnormal behavior recognition data obtained by recognition to the behavior detection model of the equipment terminal for supplementary training to form a complete closed loop, improves the detection accuracy of the behavior detection model, optimizes the behavior detection model in real time, reduces the optimization time and improves the coping ability of the behavior detection model to special scenes.
Description
Technical Field
The invention relates to the technical field of vehicle safety, in particular to a safe driving monitoring method and device for forming a data closed loop based on a cloud end.
Background
The problem of vehicle traffic safety is a worldwide problem, and the proportion of human factors is the largest (about 90%) among four factors of people, vehicles, roads and environments which form road traffic accidents. At present, the dangerous condition is judged mainly by monitoring the behavior of a driver in the safe driving of the automobile, and the algorithm model provided by an algorithm manufacturer is mainly applied to a hardware structure to monitor the behavior of the driver.
The algorithm model provided by the algorithm manufacturer in a unified way needs to be upgraded by the algorithm manufacturer in a unified way, and once the algorithm model is upgraded, the behavior of the driver can be continuously monitored by adopting the algorithm model until the next upgrade processing is carried out. Due to the mode, the identification accuracy of the algorithm model is poor, the identification error rate is high, long time is needed for iterative upgrade, and the pertinence effect of the upgraded algorithm model to a special scene is poor.
Disclosure of Invention
The invention provides a safe driving monitoring method and device for forming a data closed loop based on a cloud end, which are used for solving the defect that in the prior art, the accuracy of an algorithm is poor so that the fatigue monitoring and dangerous behavior monitoring of a driver are not in place, realizing the real-time updating of an algorithm model, reducing the time of iterative upgrade and improving the identification precision.
The invention provides a safe driving monitoring method for forming a data closed loop based on a cloud end, which comprises the following steps: receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; wherein the behavior detection model is obtained based on sample training; driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data; and inputting the abnormal behavior recognition data to the equipment terminal so as to perform supplementary training on a behavior detection model of the equipment terminal.
According to the safe driving monitoring method for forming the data closed loop based on the cloud end, the abnormal behavior detection data comprise driving images corresponding to abnormal driving behaviors obtained based on the behavior prediction model, driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data, and the method comprises the following steps: identifying a driving image in the abnormal behavior detection data to obtain an identification result corresponding to the abnormal behavior detection data; and labeling the corresponding driving image based on the recognition result to obtain abnormal behavior recognition data with a first label.
According to the safe driving monitoring method for forming the data closed loop based on the cloud end, after the abnormal behavior identification data are obtained, the method further comprises the following steps: screening the abnormal behavior identification data based on the first label to obtain a driving image of which the first label is displayed as abnormal driving behavior; generating a behavior expression analysis result of the corresponding driver based on the driving image marked as the abnormal driving behavior; and sending the behavior analysis result to a third party to which the driver belongs.
The invention also provides a safe driving monitoring method based on the cloud-end formed data closed loop, which comprises the following steps: performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to a cloud end; wherein the behavior detection model is obtained based on sample training; receiving abnormal behavior identification data which is identified by the cloud based on the abnormal behavior detection data; and performing supplementary training on the behavior detection model by using the abnormal behavior recognition data.
According to the safe driving monitoring method for forming the data closed loop based on the cloud, provided by the invention, the behavior detection is carried out on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and the method comprises the following steps: acquiring a driving image within a preset time period; inputting the acquired driving image into a behavior detection model to obtain a model prediction result output by the behavior detection model; and marking and screening corresponding driving images based on the model prediction result to obtain abnormal behavior detection data.
The invention also provides a safe driving monitoring device for forming a data closed loop based on the cloud, which comprises: the first receiving module is used for receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; wherein the behavior detection model is obtained based on sample training; the identification module is used for identifying the driving behaviors of the abnormal behavior detection data to obtain abnormal behavior identification data; and the first sending module is used for inputting the abnormal behavior identification data into the equipment end so as to perform supplementary training on a behavior detection model of the equipment end.
The invention also provides a safe driving monitoring device for forming a data closed loop based on the cloud, which comprises: the detection module is used for carrying out behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data and sending the abnormal behavior detection data to the cloud end; wherein the behavior detection model is obtained based on sample training; the second receiving module is used for receiving abnormal behavior identification data which is obtained by the cloud end based on the abnormal behavior detection data; and the control module is used for performing supplementary training on the behavior detection model by using the abnormal behavior identification data.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the processor executes the program, the steps of the safe driving monitoring method for forming the data closed loop based on the cloud end are realized.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method for safe driving monitoring based on cloud-end closed-loop formation of data as described in any of the above.
The invention also provides a computer program product, which comprises a computer program, and when the computer program is executed by a processor, the steps of the safety driving monitoring method based on cloud-based data closed loop formation are realized.
According to the safe driving monitoring method and device based on the cloud-formed data closed loop, the abnormal behavior detection data obtained through detection based on the behavior detection model are input to the cloud side by the equipment side for behavior recognition, the behavior detection model of the equipment side is subjected to supplementary training based on the abnormal behavior recognition data obtained through recognition, on the basis of forming a complete closed loop, the detection accuracy of the behavior detection model is improved, the behavior detection model is optimized in real time, the optimization time is shortened, and the response capability of the behavior detection model to special scenes is improved.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the following briefly introduces the drawings needed for the embodiments or the prior art descriptions, and obviously, the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a schematic flow chart of a safe driving monitoring method for forming a data closed loop based on a cloud end according to the present invention;
fig. 2 is a second schematic flow chart of the safety driving monitoring method for forming a data closed loop based on the cloud end according to the present invention;
fig. 3 is a third schematic flow chart of the safety driving monitoring method based on cloud-based data closed loop formation according to the present invention;
fig. 4 is a schematic structural diagram of a safe driving monitoring apparatus for forming a data closed loop based on a cloud end according to the present invention;
fig. 5 is a second schematic structural diagram of the safe driving monitoring apparatus for forming a data closed loop based on the cloud end according to the present invention;
fig. 6 is a third schematic structural diagram of the safe driving monitoring apparatus for forming a data closed loop based on the cloud end according to the present invention;
fig. 7 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in 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.
Fig. 1 shows a schematic flow chart of a safe driving monitoring method for forming a data closed loop based on a cloud, the execution subject of the method is the cloud, and the method includes:
s11, receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; the behavior detection model is obtained based on sample training;
s12, driving behavior identification is carried out on the abnormal behavior detection data to obtain abnormal behavior identification data;
and S13, inputting the abnormal behavior recognition data into the equipment terminal so as to perform supplementary training on the behavior detection model of the equipment terminal.
It should be noted that S1N in this specification does not represent the sequence of the safety driving monitoring method based on cloud-formed data closed loop, and the safety driving monitoring method based on cloud-formed data closed loop according to the present invention is specifically described below.
Step S11, receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; wherein, the behavior detection model is obtained based on sample training.
In this embodiment, the abnormal behavior detection data output by the cloud receiving device is utilized, so that secondary detection is performed on the abnormal behavior detection data subsequently, and the detection accuracy is improved. It should be noted that the device obtains abnormal behavior detection data based on the input driving image and the behavior detection model, and the abnormal behavior detection data includes: acquiring a driving image within a preset time period; inputting the acquired driving image into a behavior detection model to obtain a model prediction result output by the behavior detection model; and marking and screening the corresponding driving images based on the model prediction result to obtain abnormal behavior detection data.
It should be noted that the abnormal behavior detection data includes a model prediction result determined as an abnormal driving behavior and each frame of driving image corresponding to the model prediction result, the abnormal driving behavior includes at least one of fatigue driving, hands getting off the steering wheel, no fastening of a seat belt, eye closing, and mobile phone playing, and the specific abnormal driving behavior may be determined according to the driving behavior detected by the behavior detection model at the device side, which is not further limited herein.
In an optional embodiment, the behavior detection model is obtained by training based on samples in advance and is applied to an abnormal behavior detection algorithm model on the equipment side.
In step S12, the abnormal behavior detection data is subjected to driving behavior recognition to obtain abnormal behavior recognition data.
In this embodiment, the abnormal behavior detection data includes a driving image corresponding to the abnormal driving behavior obtained based on the behavior prediction model, and the driving behavior identification is performed on the abnormal behavior detection data to obtain the abnormal behavior identification data, including: identifying the driving image in the abnormal behavior detection data to obtain an identification result corresponding to the abnormal behavior detection data; and labeling the corresponding driving image based on the recognition result to obtain abnormal behavior recognition data with a first label.
It should be noted that, when labeling and screening the corresponding driving images based on the recognition result, labeling the corresponding driving images by using the recognition result to obtain the abnormal behavior recognition data with the first tag. In addition, the first label comprises an abnormal driving behavior and a normal driving behavior, when the corresponding driving image is labeled by using the identification result, the corresponding driving image can be directly labeled as the abnormal driving behavior or the normal driving behavior, and the corresponding abnormal behavior identification data is the driving image labeled as the abnormal driving behavior and/or the driving image labeled as the normal driving behavior. The driving image in the abnormal behavior detection data is identified, so that the abnormal behavior detection data sent by the equipment terminal can be detected for the second time, and the accuracy of the detection result is improved.
In an optional embodiment, after obtaining the abnormal behavior recognition data, the method further includes: screening the abnormal behavior identification data based on the first label to obtain a driving image of which the first label is displayed as the abnormal driving behavior; generating a behavioral expression analysis result of the corresponding driver based on the driving image labeled as the abnormal driving behavior; and sending the behavior analysis result to a third party to which the driver belongs.
In order to facilitate matching of the driving image labeled as the abnormal driving behavior with the corresponding driver, after the device side acquires the driving image within the preset time period, the method further comprises the following steps: the equipment terminal carries out face recognition on the acquired driving image so as to determine the identity of a corresponding driver, and sends a face recognition result to the cloud; or after the abnormal behavior detection data sent by the cloud receiving device end, performing face recognition on the driving image in the abnormal behavior detection data to determine the identity of the corresponding driver.
It should be noted that when the behavior analysis result of the corresponding driver is generated based on the driving image labeled as the abnormal driving behavior, the behavior analysis result of the corresponding driver is generated by combining the identity of the corresponding driver, so as to correspondingly display whether the abnormal driving behavior exists in the driving period of the corresponding driver, thereby facilitating the evaluation of the driving behavior of the driver by the third party to which the corresponding driver belongs according to the behavior analysis result, and further facilitating the corresponding reward and punishment of the corresponding driver by the third party.
And step S13, inputting the abnormal behavior recognition data into the equipment terminal so as to perform supplementary training on the behavior detection model of the equipment terminal.
In this embodiment, after the cloud obtains the abnormal behavior recognition data, the abnormal behavior recognition data is input to the device side, so that the device side can perform supplementary training on the behavior detection model by using the abnormal behavior recognition data, and predict a newly acquired driving image in a preset time period by using the behavior detection model after the supplementary training is completed, so that the detection accuracy of the behavior detection model is improved, the behavior detection model is optimized in time, the optimization time is shortened, and the response capability of the behavior detection model to a special scene is improved.
The abnormal behavior identification data is obtained by detecting the abnormal behavior detection data for the second time, after the abnormal behavior detection data and the abnormal behavior identification data are input to the equipment terminal, the driving image contained in the abnormal behavior detection data is input to the behavior detection model as training data to obtain a model training result output by the behavior detection model, and the model training result is compared with the first label corresponding to the abnormal behavior identification result based on the model training result to judge whether the training is finished.
Fig. 2 is a schematic flow chart of a safe driving monitoring method for forming a data closed loop based on a cloud, the method including:
s21, performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to a cloud end; wherein the behavior detection model is obtained based on sample training;
s22, receiving abnormal behavior identification data which is obtained by the cloud end based on the abnormal behavior detection data;
and S23, performing supplementary training on the behavior detection model by using the abnormal behavior recognition data.
It should be noted that S2N in this specification does not represent the sequence of the safety driving monitoring method based on cloud-formed data closed loop, and the safety driving monitoring method based on cloud-formed data closed loop according to the present invention is specifically described below.
Step S21, performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to the cloud; wherein, the behavior detection model is obtained based on sample training.
In this embodiment, the behavior detection performed on the received driving image based on the behavior detection model to obtain abnormal behavior detection data includes: acquiring a driving image within a preset time period; inputting the acquired driving image into a behavior detection model to obtain a model prediction result output by the behavior detection model; and marking and screening the corresponding driving images based on the model prediction result to obtain abnormal behavior detection data.
It should be noted that, labeling and screening the corresponding driving image based on the model prediction result to obtain abnormal behavior detection data includes: labeling the corresponding driving image based on the model prediction result to obtain a driving image with a second label; wherein the second label comprises abnormal driving behavior and normal driving behavior; and screening the driving image based on the second label to obtain the driving image with the second label displayed as the abnormal driving behavior, namely the abnormal behavior detection data.
In addition, the abnormal driving behavior includes at least one of fatigue driving, hands being detached from a steering wheel, unfastening a safety belt, closing eyes, and playing a mobile phone, and the specific abnormal driving behavior may be determined according to a driving behavior detected by a behavior detection model on the device side, which is not further limited herein.
In an alternative embodiment, acquiring the driving image within the preset time period includes: acquiring a video stream of a driving area; and extracting the driving images in a preset time period based on the video stream. Specifically, based on a video stream, extracting a driving image within a preset time period includes: the method includes the steps that driving images of a certain number of frames before and after a current frame driving image are collected based on a video stream, for example, driving images of preset number of frames before and after the current frame driving image are collected to avoid the situation that a detection result is poor due to the fact that subsequent judgment is conducted based on a single frame driving image, the number of the obtained specific driving images can be set according to actual detection needs, further limitation is not needed, and for example, the driving images of five frames before and after the current frame driving image can be obtained. In other embodiments, acquiring the driving image within the preset time period includes: the method comprises the steps of obtaining a plurality of frames of continuously shot driving images in a preset time period.
The driving image is an image captured based on a driving area in the vehicle. The vehicle may be a vehicle, a ship, an airplane, or other vehicles for carrying people or goods, wherein the vehicle may be a private car or an operating vehicle, such as a shared automobile, a network appointment car, a taxi, a bus, a school bus, a truck, a passenger car, a train, a subway, a rail electric car, and the like.
In an optional embodiment, a high-resolution infrared camera can be used to acquire the video stream, the infrared camera is arranged in the driving area of the vehicle owner to shoot the driving area, and an infrared lamp can be arranged in the corresponding vehicle so as to clearly acquire the driving image at night or under the condition of poor illumination condition. In other embodiments, the driving image collecting device may adopt at least one of a DMS camera, an OMS camera, a video recorder, and an electronic device with a camera, and the electronic device may be a mobile terminal device such as a mobile terminal, a computer, a camera, a tablet computer, and the like.
In an optional embodiment, the behavior detection model is obtained by training based on samples in advance and is applied to an abnormal behavior detection algorithm model on the equipment side.
In an optional embodiment, after the abnormal behavior detection data is sent to the cloud, the cloud performs driving behavior recognition based on the abnormal behavior detection data to obtain abnormal behavior recognition data, and returns the abnormal behavior recognition data to the device side. For the specific identification of the abnormal behavior detection data by the cloud, reference may be made to the foregoing method embodiment, which is not described herein again.
Step S22, receiving abnormal behavior identification data identified by the cloud based on the abnormal behavior detection data.
Step S23, the behavior detection model is subjected to supplementary training using the abnormal behavior recognition data.
In this embodiment, the performing supplementary training on the behavior detection model by using the abnormal behavior recognition data includes: inputting the abnormal behavior recognition data into the behavior detection model to obtain a model training result output by the behavior detection model; and comparing the model training result with the first label corresponding to the abnormal behavior identification data to judge whether the training is finished.
In an optional embodiment, after the performing the supplementary training on the behavior detection model by using the abnormal behavior recognition data and the abnormal behavior detection data, the method further includes: safety driving monitoring based on a post-training behavior detection model, in other words, i.e. safety driving monitoring
Fig. 3 is a schematic flow chart of a safe driving monitoring method for forming a data closed loop based on a cloud, the method including:
s31, the equipment side performs behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data and sends the abnormal behavior detection data to the cloud; wherein the behavior detection model is obtained based on sample training;
s32, the cloud end carries out driving behavior recognition on the received abnormal behavior detection data to obtain abnormal behavior recognition data, and the abnormal behavior recognition data is returned to the equipment end;
and S33, the equipment side carries out supplementary training on the behavior detection model by using the received abnormal behavior identification data.
It should be noted that S3N in this specification does not represent the sequence of the safety driving monitoring method based on cloud-formed data closed loop, and the safety driving monitoring method based on cloud-formed data closed loop according to the present invention is specifically described below.
S31, the equipment side performs behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data and sends the abnormal behavior detection data to the cloud; wherein the behavior detection model is obtained based on sample training.
In this embodiment, the behavior detection of the device on the basis of the received driving image by the behavior detection model to obtain abnormal behavior detection data includes: acquiring a driving image in a preset time period; inputting the acquired driving image into a behavior detection model to obtain a model prediction result output by the behavior detection model; and marking and screening the corresponding driving images based on the model prediction result to obtain abnormal behavior detection data. Reference is made to the foregoing description for specific embodiments, which are not repeated herein.
And S32, the cloud end carries out driving behavior recognition on the received abnormal behavior detection data to obtain abnormal behavior recognition data, and the abnormal behavior recognition data is returned to the equipment end.
In this embodiment, the cloud end performs driving behavior identification on the received abnormal behavior detection data to obtain abnormal behavior identification data, including: identifying a driving image within the abnormal behavior detection data; obtaining an identification result corresponding to the abnormal behavior detection data; and labeling the corresponding driving image based on the recognition result to obtain abnormal behavior recognition data with a first label.
In an optional embodiment, after obtaining the abnormal behavior recognition data at the cloud, the method further includes: screening the abnormal behavior identification data based on the first label to obtain a driving image of which the first label displays abnormal driving behaviors; generating a behavioral expression analysis result of the corresponding driver based on the driving image labeled as the abnormal driving behavior; and sending the behavior analysis result to a third party to which the driver belongs.
And S33, the equipment side carries out supplementary training on the behavior detection model by using the received abnormal behavior identification data.
In this embodiment, the performing, by the device side, supplementary training on the behavior detection model by using the received abnormal behavior identification data includes: inputting the abnormal behavior recognition data into the behavior detection model to obtain a model training result output by the behavior detection model; and comparing the model training result with the first label corresponding to the abnormal behavior identification data to judge whether the training is finished.
In summary, the device end inputs abnormal behavior detection data obtained through detection based on the behavior detection model to the cloud end for behavior recognition, and performs supplementary training on the behavior detection model of the device end based on the abnormal behavior recognition data obtained through recognition, so that on the basis of forming a complete closed loop, the detection accuracy of the behavior detection model is improved, the behavior detection model is optimized in real time, the optimization time is reduced, and the response capability of the behavior detection model to a special scene is improved.
The safety driving monitoring device based on the cloud-formed data closed loop provided by the invention is described below, and the safety driving monitoring device based on the cloud-formed data closed loop and the safety driving monitoring method based on the cloud-formed data closed loop described above can be referred to correspondingly.
Fig. 4 is a schematic structural diagram of a safe driving monitoring device forming a data closed loop based on a cloud, the device being a cloud, the device including:
a first receiving module 41, configured to receive abnormal behavior detection data obtained by the device based on the input driving image and the behavior detection model; wherein the behavior detection model is obtained based on sample training;
the identification module 42 is used for identifying the driving behaviors of the abnormal behavior detection data to obtain abnormal behavior identification data;
the first sending module 43 inputs the abnormal behavior recognition data to the device side to perform supplementary training on the behavior detection model of the device side.
In this embodiment, the first receiving module 41 includes: the device comprises a first receiving unit, a second receiving unit and a control unit, wherein the first receiving unit is used for receiving abnormal behavior detection data obtained by the device based on an input driving image and a behavior detection model; wherein, the behavior detection model is obtained based on sample training.
An identification module 42 comprising: the first detection unit is used for identifying the driving image in the abnormal behavior detection data to obtain an identification result corresponding to the abnormal behavior detection data; and the first labeling unit labels the corresponding driving image based on the identification result to obtain abnormal behavior identification data with a first label. It should be noted that, when the corresponding driving images are labeled and screened based on the identification result, the corresponding driving images are labeled by using the identification result, so as to obtain the abnormal behavior identification data with the first label. In addition, the first label comprises an abnormal driving behavior and a normal driving behavior, when the corresponding driving image is labeled by using the identification result, the corresponding driving image can be directly labeled as the abnormal driving behavior or the normal driving behavior, and the corresponding abnormal behavior identification data is the driving image labeled as the abnormal driving behavior and/or the driving image labeled as the normal driving behavior. The driving image in the abnormal behavior detection data is identified, so that the abnormal behavior detection data sent by the equipment terminal is detected for the second time, and the accuracy of the detection result is improved.
In an optional embodiment, the apparatus further comprises an evaluation module for generating a performance analysis result based on the abnormal behavior recognition data and sending the performance analysis result to a third party to which the driver belongs. . In particular, the evaluation module comprises: the first screening unit is used for screening the abnormal behavior identification data based on the first label to obtain a driving image of which the first label is displayed as the abnormal driving behavior; a first labeling unit which generates a behavior expression analysis result of a corresponding driver based on the driving image labeled as the abnormal driving behavior; and the data feedback unit is used for sending the behavior analysis result to a third party to which the driver belongs.
The first sending module 43, which includes a data sending unit, inputs the abnormal behavior recognition data to the device side, so as to perform supplementary training on the behavior detection model of the device side.
Fig. 5 is a schematic structural diagram of a safe driving monitoring apparatus for forming a data closed loop based on a cloud, the apparatus being a device side, the apparatus including:
the detection module 51 is used for performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data and sending the abnormal behavior detection data to the cloud end; the behavior detection model is obtained based on sample training;
a second receiving module 52, configured to receive abnormal behavior identification data obtained by the cloud based on the abnormal behavior detection data;
the control module 53 performs supplementary training on the behavior detection model using the abnormal behavior recognition data.
In this embodiment, the detection module 51 includes: the image acquisition unit is used for acquiring a driving image in a preset time period; the prediction unit is used for inputting the acquired driving image into the behavior detection model to obtain a model prediction result output by the behavior detection model; and the data acquisition unit is used for marking and screening the corresponding driving images based on the model prediction result to obtain abnormal behavior detection data.
Specifically, the data acquisition unit includes: the labeling subunit labels the corresponding driving image based on the model prediction result to obtain a driving image with a second label; wherein the second label comprises abnormal driving behavior and normal driving behavior; and the screening subunit screens the driving image based on the second label to obtain a driving image with the second label displayed as abnormal driving behavior, namely abnormal behavior detection data.
In addition, the abnormal driving behavior includes at least one of fatigue driving, hands being detached from a steering wheel, unfastening a safety belt, closing eyes, and playing a mobile phone, and the specific abnormal driving behavior may be determined according to a driving behavior detected by a behavior detection model on the device side, which is not further limited herein.
An image acquisition unit comprising: the camera shooting subunit acquires a video stream of the driving area; and an image extraction subunit which extracts the driving image within a preset time period based on the video stream. Specifically, the image extracting subunit acquires, based on the video stream, driving images of a certain number of frames before and after the current frame driving image, for example, driving images of five frames before and after the current frame driving image, so as to avoid a situation that a subsequent determination result and a detection result based on a single frame driving image are poor, and the number of the acquired specific driving images can be set according to an actual detection requirement, which is not further limited herein. In other embodiments, an image acquisition unit includes: and the single-frame image acquisition sub-unit is used for acquiring a plurality of frames of continuously shot driving images within a preset time period.
The second receiving module 52 includes a data receiving unit, which receives abnormal behavior identification data identified by the cloud based on the abnormal behavior detection data.
A control module 53 comprising: the training unit is used for inputting the abnormal behavior recognition data into the behavior detection model to obtain a model training result output by the behavior detection model; and the stopping unit is used for comparing the model training result with the first label corresponding to the abnormal behavior identification data so as to judge whether the training is finished.
Fig. 6 is a schematic structural diagram of a safe driving monitoring apparatus forming a data closed loop based on a cloud, the apparatus including a device end 61 and a cloud end 62, wherein:
the device end 61 performs behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sends the abnormal behavior detection data to the cloud end 62; wherein the behavior detection model is obtained based on sample training;
the cloud 62 performs driving behavior recognition on the received abnormal behavior detection data to obtain abnormal behavior recognition data, and returns the abnormal behavior recognition data to the device end 61;
the device side 61 performs supplementary training on the behavior detection model by using the received abnormal behavior recognition data.
It should be noted that, the specific structures of the device end 61 and the cloud end 62 may refer to the device end apparatus embodiment and the cloud end apparatus embodiment described above, and details are not described herein.
Fig. 7 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 7: a processor (processor)71, a communication Interface (Communications Interface)72, a memory (memory)73 and a communication bus 74, wherein the processor 71, the communication Interface 72 and the memory 73 are communicated with each other via the communication bus 74. The processor 71 may call logic instructions in the memory 73 to perform a cloud-based data closed-loop safe driving monitoring method, which includes: receiving abnormal behavior detection data obtained by the equipment terminal based on the input driving image and the behavior detection model; the behavior detection model is obtained based on sample training; driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data; inputting the abnormal behavior recognition data to the equipment end to perform supplementary training on a behavior detection model of the equipment end; or performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to the cloud; the behavior detection model is obtained based on sample training; receiving abnormal behavior identification data which is obtained by the cloud terminal based on the abnormal behavior detection data; and performing supplementary training on the behavior detection model by using the abnormal behavior recognition data.
Further, the logic instructions in the memory 73 may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a stand-alone product. 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 another aspect, the present invention further provides a computer program product, where the computer program product includes a computer program that can be stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer is capable of executing the method for cloud-based data closed-loop secure driving monitoring provided by the above methods, where the method includes: receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; the behavior detection model is obtained based on sample training; driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data; inputting the abnormal behavior identification data to the equipment end to perform supplementary training on a behavior detection model of the equipment end; or, performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to the cloud; the behavior detection model is obtained based on sample training; receiving abnormal behavior identification data which is obtained by the cloud end based on the abnormal behavior detection data; and performing supplementary training on the behavior detection model by using the abnormal behavior recognition data.
In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented to, when executed by a processor, perform the method for monitoring safe driving based on cloud-formed data closed loop provided by the foregoing methods, where the method includes: receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; the behavior detection model is obtained based on sample training; driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data; inputting the abnormal behavior identification data to the equipment end to perform supplementary training on a behavior detection model of the equipment end; or, performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to the cloud; the behavior detection model is obtained based on sample training; receiving abnormal behavior identification data which is obtained by the cloud end based on the abnormal behavior detection data; and performing supplementary training on the behavior detection model by using the abnormal behavior recognition data.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate components may or may not be physically separate, and components 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 can be selected according to actual needs to achieve the purpose of the solution of the embodiment. One of ordinary skill in the art can understand and implement the present invention without any 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 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 method 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 should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may 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.
Claims (10)
1. A safe driving monitoring method based on cloud-end formed data closed loop is characterized by comprising the following steps:
receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; wherein the behavior detection model is obtained based on sample training;
driving behavior recognition is carried out on the abnormal behavior detection data to obtain abnormal behavior recognition data;
and inputting the abnormal behavior recognition data to the equipment terminal so as to perform supplementary training on a behavior detection model of the equipment terminal.
2. The cloud-based data closed-loop safe driving monitoring method of claim 1, wherein the abnormal behavior detection data comprises a driving image corresponding to abnormal driving behavior obtained based on the behavior prediction model, and the driving behavior recognition is performed on the abnormal behavior detection data to obtain abnormal behavior recognition data, and the method comprises:
identifying a driving image in the abnormal behavior detection data to obtain an identification result corresponding to the abnormal behavior detection data;
and labeling the corresponding driving image based on the identification result to obtain abnormal behavior identification data with a first label.
3. The cloud-based data closed-loop safe driving monitoring method of claim 2, wherein after obtaining the abnormal behavior recognition data, the method further comprises:
screening the abnormal behavior identification data based on the first label to obtain a driving image of which the first label is displayed as abnormal driving behavior;
generating a behavioral expression analysis result of the corresponding driver based on the driving image marked as the abnormal driving behavior;
and sending the behavior analysis result to a third party to which the driver belongs.
4. A safe driving monitoring method based on cloud-end formed data closed loop is characterized by comprising the following steps:
performing behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data, and sending the abnormal behavior detection data to a cloud end; wherein the behavior detection model is obtained based on sample training;
receiving abnormal behavior identification data which is identified by the cloud based on the abnormal behavior detection data;
and performing supplementary training on the behavior detection model by using the abnormal behavior recognition data.
5. The cloud-based data closed-loop safe driving monitoring method of claim 4, wherein the behavior detection of the received driving image based on the behavior detection model to obtain abnormal behavior detection data comprises:
acquiring a driving image within a preset time period;
inputting the acquired driving image into a behavior detection model to obtain a model prediction result output by the behavior detection model;
and marking and screening corresponding driving images based on the model prediction result to obtain abnormal behavior detection data.
6. The utility model provides a safe driving monitoring device based on high in clouds forms data closed loop which characterized in that includes:
the first receiving module is used for receiving abnormal behavior detection data obtained by the equipment based on the input driving image and the behavior detection model; wherein the behavior detection model is obtained based on sample training;
the identification module is used for identifying the driving behaviors of the abnormal behavior detection data to obtain abnormal behavior identification data;
and the first sending module is used for inputting the abnormal behavior identification data to the equipment terminal so as to perform supplementary training on a behavior detection model of the equipment terminal.
7. The utility model provides a safe driving monitoring device based on high in clouds forms data closed loop which characterized in that includes:
the detection module is used for carrying out behavior detection on the received driving image based on the behavior detection model to obtain abnormal behavior detection data and sending the abnormal behavior detection data to the cloud end; wherein the behavior detection model is obtained based on sample training;
the second receiving module is used for receiving abnormal behavior identification data which is obtained by the cloud end based on the abnormal behavior detection data;
and the control module is used for performing supplementary training on the behavior detection model by using the abnormal behavior identification data.
8. 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 steps of the cloud-based data closed-loop safe driving monitoring method according to any one of claims 1 to 3 when executing the program; or implementing the steps of the method for safe driving monitoring based on cloud-based data closed loop formation according to any one of claims 4 to 5.
9. A non-transitory computer readable storage medium, on which a computer program is stored, wherein the computer program, when being executed by a processor, implements the steps of the cloud-based safe driving monitoring method for forming a data closed loop according to any one of claims 1 to 3; or implementing the steps of the method for safe driving monitoring based on cloud-based data closed loop formation according to any one of claims 4 to 5.
10. A computer program product comprising a computer program, wherein the computer program when executed by a processor implements the steps of the cloud-based closed-loop data-based safe driving monitoring method of any one of claims 1 to 3; or implementing the steps of the method for safe driving monitoring based on cloud-based data closed loop formation according to any one of claims 4 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111016009.8A CN113887297A (en) | 2021-08-31 | 2021-08-31 | Safe driving monitoring method and device for forming data closed loop based on cloud |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111016009.8A CN113887297A (en) | 2021-08-31 | 2021-08-31 | Safe driving monitoring method and device for forming data closed loop based on cloud |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113887297A true CN113887297A (en) | 2022-01-04 |
Family
ID=79011499
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111016009.8A Pending CN113887297A (en) | 2021-08-31 | 2021-08-31 | Safe driving monitoring method and device for forming data closed loop based on cloud |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113887297A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115439954A (en) * | 2022-08-29 | 2022-12-06 | 上海寻序人工智能科技有限公司 | Data closed-loop method based on cloud large model |
-
2021
- 2021-08-31 CN CN202111016009.8A patent/CN113887297A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115439954A (en) * | 2022-08-29 | 2022-12-06 | 上海寻序人工智能科技有限公司 | Data closed-loop method based on cloud large model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3777122B1 (en) | Image processing method and apparatus | |
| CN111274881B (en) | Driving safety monitoring method and device, computer equipment and storage medium | |
| CN109902575B (en) | Anti-walking method and device based on unmanned vehicle and related equipment | |
| CN110889351B (en) | Video detection method, device, terminal equipment and readable storage medium | |
| US9129159B2 (en) | Vehicle headlight state monitoring methods, systems and processor-readable media | |
| CN107967806A (en) | Vehicle fake-license detection method, device, readable storage medium storing program for executing and electronic equipment | |
| CN111291812B (en) | Method and device for acquiring attribute category, storage medium and electronic device | |
| US20210327040A1 (en) | Model training method and system for automatically determining damage level of each of vehicle parts on basis of deep learning | |
| US20150279036A1 (en) | Side window detection in near-infrared images utilizing machine learning | |
| US11250279B2 (en) | Generative adversarial network models for small roadway object detection | |
| CN106292625B (en) | Vehicle fault online detection system and method | |
| CN109145843A (en) | A kind of full vehicle information identification system of bayonet high definition camera and method | |
| US11482012B2 (en) | Method for driving assistance and mobile device using the method | |
| CN112699754B (en) | Signal lamp identification method, device, equipment and storage medium | |
| CN106779089A (en) | Method of testing and device | |
| US11120308B2 (en) | Vehicle damage detection method based on image analysis, electronic device and storage medium | |
| CN111985373A (en) | Safety early warning method and device based on traffic intersection identification and electronic equipment | |
| CN109859085A (en) | Safe early warning method and Related product | |
| CN113887297A (en) | Safe driving monitoring method and device for forming data closed loop based on cloud | |
| CN110443119B (en) | Method and device for identifying state of goods in carriage | |
| CN111241918B (en) | Vehicle tracking prevention method and system based on face recognition | |
| CN110853364B (en) | Data monitoring method and device | |
| CN110188645B (en) | Face detection method and device for vehicle-mounted scene, vehicle and storage medium | |
| CN113378653A (en) | Method and device for detecting in-vehicle behavior of driver | |
| CN113537117A (en) | Vehicle-mounted legacy monitoring and alarming method and 5G system thereof |
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 |