CN113569733A - Safe driving recommendation method based on saccade of driver in urban environment - Google Patents
Safe driving recommendation method based on saccade of driver in urban environment Download PDFInfo
- Publication number
- CN113569733A CN113569733A CN202110854240.8A CN202110854240A CN113569733A CN 113569733 A CN113569733 A CN 113569733A CN 202110854240 A CN202110854240 A CN 202110854240A CN 113569733 A CN113569733 A CN 113569733A
- Authority
- CN
- China
- Prior art keywords
- driving
- subject
- eye
- saccade
- model
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration by the use of local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Abstract
The invention discloses a safe driving recommendation method in an urban environment based on saccades of drivers, which comprises the following steps: collecting a driving video of a subject, obtaining a fixation point of the subject and a model, and obtaining key characteristics by using a classical attention model; constructing a neural network, and selecting the most appropriate fixation point; determining a finally recommended safe driving scheme by optimizing the fixation point; according to the invention, under a crowded road environment, by comparing the differences of the scanning tracks between the subjects and the model, the accuracy of scanning can be improved, the safety is met, a safe driving scheme is recommended, the safe driving is realized, and a reference is provided for the future completely autonomous driving.
Description
Technical Field
The invention relates to a driver attention analysis and modeling technology, in particular to a safe driving recommendation method in an urban environment based on eye saccades of drivers.
Background
In recent decades, computational modeling of human visual attention has received widespread attention. In modern advanced industrial applications, it has proven to be very similar to predicting human visual attention. However, it is controversial whether or not the solution of predicting eye saccade positions based on eye saccade positions of human Visual Attention (Visual Saccades) and a Computational Visual Attention Model (Saccades Predicted by Computational Visual Attention Model) is more reliable and actually contributes to actual driving.
Considering that the biological inspiration of a driving-related attention model can be obtained from a skilled driver under complex driving conditions, in which the attention of the driver alternately gazes through eye jumps to drive safely, thereby continuously pointing to various remarkable and information-rich visual stimuli, an eye jump recommendation strategy for improving driving safety under crowded road environments, particularly when the vision of the driver is often affected by the visual crowding, is proposed herein.
Disclosure of Invention
In view of the above, the present invention provides a recommendation method for safe driving in an urban environment based on saccades of drivers, so as to solve the above technical problems.
In order to achieve the purpose, the invention provides the following technical scheme:
a safe driving recommendation method in an urban environment based on a driver eye saccade comprises the following steps:
step 1: acquiring and preprocessing an original driving video, acquiring the original driving video of a virtual and real city driving scene, and preprocessing the original driving video;
step 2: obtaining an eye-movement track and a fixation point of a subject and a model, calling the subject with driving experience to participate in an eye-movement experiment, obtaining the eye-movement track common to the subject, and obtaining the eye-movement track common to the subject through an eye-movement equipment; selecting an attention model to predict a fixation point and a corresponding glancing track of the preprocessed driving video;
and step 3: preliminarily selecting a glance trajectory scheme, and preliminarily selecting an optimal scheme of the glance trajectory by comparing the fixation points of the subject and the model with the glance differences;
and 4, step 4: extracting key features related to saccades in the preprocessed original driving video;
and 5: constructing a neural network model, selecting a most reasonable fixation point, constructing the neural network model, and selecting the most reasonable fixation point by referring to an overall safe driving scheme to meet the requirement of safe driving;
step 6: and (5) optimizing the fixation point in the step (5), determining a final recommended safe driving scheme, optimizing the fixation point selected in the step (5) in one step, improving driving comfort, and enabling the subject to participate in the scheme through subjective evaluation to finally promote a target mark of safe driving.
Further, the acquiring of the original driving video of the virtual and real city driving scenes comprises: half of the original driving video is shot under the field condition, the other half is shot under the virtual scene, the original driving video adopts a digital camera arranged on the windshield of the automobile, and in addition, a high-definition camera is used for shooting, and the frames per second are 30.
Further, the digital video camera installed on the windshield of the automobile extracts 6 high-definition color video clips from 10 driving environments of expressways and urban roads in real scenes and virtual scenes.
Further, the preprocessing of the original driving video comprises performing color thresholding processing and gaussian filtering processing on the original driving video.
Further, the subjects include at least 35 subjects, all of whom have a driving experience of at least one year or a driving history of 10000 km or more, and all of whom have normal vision or normal correction and normal color vision.
Further, step 2 specifically includes: the step 2 specifically comprises the following steps: the attention model uses a classical attention model, and 2 different saccadic eye scan schemes of the subject's gaze point and the model predicted gaze point are obtained by analyzing the pre-processed original driving video using the classical attention model.
Further, step 3 includes 4 key features associated with saccades, including: four characteristics of the difference of the significant markers of the two adjacent frames designated by the driving expert, the distance difference between the eye movement scanning viewpoint of the subject and the eye movement scanning viewpoint of the model between the two adjacent frames, the distance difference between the two eye movement scanning schemes of the same frame and the distance difference between the eye movement scanning viewpoint of the subject and the eye movement scanning viewpoint of the model between the forward and backward videos and the difference between the optimal significant marker and the suboptimal significant marker are analyzed.
Further, the key feature extraction method is as follows:
step 4.1: obtaining saccade data of the subject and the model through saccade instrument and model prediction;
step 4.2: and extracting the saccades by adopting wavelet analysis and an approximate entropy algorithm.
Further, step 5 specifically includes: and selecting a forward propagation network training, prejudging and verifying a sweep point in the residual preprocessed original driving video according to the key characteristics.
Further, step 6 specifically includes: and after the recommended point scheme in the step 5 is obtained, selecting to further optimize the fixation point according to the visual comfort and safety, driving the subject according to the recommended fixation point, and then participating in judgment and scoring.
The technical scheme can show that the invention has the advantages that:
compared with the prior art, under a crowded road environment, especially when the vision of a driver is often influenced by visual crowding, the method selects a more reasonable and more optimal scheme by analyzing key features and a glancing track of the driver, which are predicted by a subject and a model and are related to eye saccades, so as to achieve safe driving and provide reference for realizing full automatic driving in the future.
In addition to the objects, features and advantages described above, other objects, features and advantages of the present invention are also provided. The present invention will be described in further detail below with reference to the drawings.
Drawings
In the drawings:
fig. 1 is a schematic flow chart of a method for recommending safe driving in an urban environment based on saccades of drivers according to the present invention.
Fig. 2 is a city driving scene shot using a webcam installed on a mini SUV (modern ix25) according to an embodiment of the present invention.
Fig. 3 is a highway scene shot using a webcam installed on a mini SUV (modern ix25) according to an embodiment of the present invention.
Fig. 4 is a driving scene from an urban road photographed from a UE4 (fantasy engine 4) on a Visual Studio 2019 platform according to an embodiment of the present invention.
Fig. 5 is a driving scene from an expressway captured from a UE4 (ghost engine 4) on a Visual Studio 2019 platform according to an embodiment of the present invention.
FIG. 6 is a static eye saccade scatter plot of a driving video clip after application of T-SNE in accordance with an embodiment of the present invention.
FIG. 7 is a dynamic eye saccade scatter plot of a driving video clip after application of T-SNE in accordance with an embodiment of the present invention.
FIG. 8 is a diagram of eye saccade clusters for highway driving tasks in an express environment in accordance with an embodiment of the present invention.
FIG. 9 is a diagram of eye saccade clusters under a driving task of an urban environment road according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be 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.
Referring to fig. 1 to 9, the invention discloses a recommendation method for safe driving in an urban environment based on saccades of drivers, which comprises the following steps:
step 1: the acquisition and pre-processing of the original driving video,
calling subjects with driving experience to participate in eye movement experiments, carrying out original driving video acquisition on virtual and real urban driving scenes, and preprocessing the original driving videos;
step 2: obtaining the saccade locus and the fixation point of the subject and the model,
obtaining a subject's common saccadic eye trajectory by a saccadic eye device; selecting a representative attention model to predict a fixation point and a corresponding glancing track of the preprocessed original driving video;
and step 3: a glance trajectory scenario was initially selected,
preliminarily selecting an optimal scheme of a saccade track by comparing the differences of the fixation points and saccades of the subject and the model;
and 4, step 4: extracting key features related to saccadic eye movements in the preprocessed original driving video;
and 5: constructing a neural network model, selecting the most reasonable fixation point,
constructing a neural network model, and selecting the most reasonable fixation point by referring to an overall safe driving scheme;
step 6: the point of regard of step 5 is optimized, the final safe driving scheme is determined,
and (5) optimizing the selected fixation point in the step 5, improving driving comfort, and enabling the subject to participate in the subjective evaluation scheme, thereby finally promoting the goal of safe driving.
Specifically, in the implementation process of the invention, a small-sized passenger car (below 7 seats) and a high-definition camera are adopted to record the external driving scene.
S1: as much as possible, comprehensive and rich virtual and real driving scenes are collected for collection and pretreatment, such as natural scenes of expressways, urban roads, expressways, rainy days, foggy days and the like;
s2: obtaining, by an eye tracker device, saccade data of the subject, while obtaining predicted saccade data by an attention model;
s3: by analyzing the subject's saccades and predicting differences in saccades;
s4: extracting multi-dimensional key features of the preprocessed original driving video;
s5: training a neural network, judging a fixation point under different driving scenes, referring to the opinions (ground truth) of a driving expert to provide a scheme for recommending safe driving, and verifying the reasonability of the scheme;
s6: the selection of the fixation point is optimized, the comfort is improved to further improve the safety, and finally, a safe driving scheme is determined and recommended.
As shown in fig. 2 to 5, a small bus (7 seats or less) with a wide foreground view is required for the experiment when data is collected. Meanwhile, the original driving video database is to collect all driving tasks encountered by the driver in real life as much as possible, such as: virtual and real city driving scenes include driving scenes of complex, simple expressways, urban roads, and emphasizing different weather conditions. All original driving videos need to consider both left (e.g. uk or australia) and right (e.g. china or usa) rudder scenarios. Half of the original driving video is shot in bad weather conditions, i.e., rain, fog, snow, strong wind, night, etc., and the other half is shot in sunny days. 6 high-definition color video clips were extracted from 10 different driving environments and recorded with a digital video camera mounted on the windshield of the vehicle. Additionally captured using a panasonic HX-DC3 high definition camera with a resolution of 640 x 480 pixels at 30 frames per second. High definition cameras are fixed on tripods to ensure good image quality. .
Specifically, fig. 2 is city driving picture information collected by a webcam installed on a mini SUV, fig. 3 is highway picture information collected by a webcam installed on a mini SUV, fig. 4 is a driving scene of a city road photographed by a UE4 on a virtual platform, and fig. 5 is a driving scene of a highway photographed by a UE4 on a virtual platform. The driving pictures in different scenes are collected so as to ensure the universality of the collected driving pictures.
At least 35 subjects, between 21 and 42 years of age, and on average 32.4 ± 0.42 years of age (mean ± sem) adults (7 females, 28 males) were required to voluntarily participate in this study. All participants had at least one year of driving experience or had driving records of 10000 km and more. All subjects had normal vision or corrected to normal, and color vision was normal. Meanwhile, a classical attention model is adopted to analyze the original driving video to obtain 2 different schemes, namely a subject fixation point and a model predicted fixation point. .
The preprocessing of the original driving video comprises the steps of carrying out color thresholding processing and Gaussian filtering processing on the original driving video.
Specifically, the device for collecting the saccade of the subject is an eye movement instrument which adopts an eye movement instrument preferred under an eye movement platform 2.0.
The attention model adopts a classical attention model, and the subject gaze point and the model predicted gaze point are obtained by analyzing the preprocessed original driving video by adopting the classical attention model.
Four key saccadic characteristics are analyzed and extracted, namely four characteristics of (1) the difference of a significant marker between two adjacent frames is designated by an analysis expert driver, (2) the distance difference between the eye movement viewpoint of a subject and the eye movement viewpoint of a model between the two adjacent frames, (3) the distance difference between two saccadic schemes between the same frame and the distance difference between the eye movement viewpoint of the subject and the eye movement viewpoint of the model between a forward and backward video, (4) and the difference of an optimal significant marker and a suboptimal significant marker.
Specifically, the key feature extraction approach related to saccades is as follows:
step a: obtaining saccade data of the subject by a saccade instrument;
step b: wavelet analysis and approximate entropy algorithm are adopted to extract key features related to saccades. And selecting a forward propagation network training, prejudging and verifying a sweep point in the driving video after the residual preprocessing. The visual effect of high-dimensional classification of bet point-of-view data for virtual and real driving scenarios is shown in fig. 6-7. .
Specifically, as shown in FIG. 6, the present embodiment captures a saccade scatter plot of a driving video clip after applying T-SNE. Wherein: blue (1) indicates saccadic eye scatter on other static visual stimuli, red (2) indicates saccadic eye scatter on traffic lights, and green (3) indicates saccadic eye scatter on traffic signs. As shown in fig. 7, the present embodiment applies yet another saccade scattergram of driving video acquired after T-SNE. Green (4) for saccadic astigmatism on pedestrian dynamic stimuli, (5) blue for saccadic astigmatism on other dynamic visual stimuli, and (6) red for saccadic astigmatism on moving vehicle visual stimuli. By collecting static and dynamic saccade scatter diagrams, more accurate basis is provided for selecting safe driving schemes.
And after the scheme of the fixation point is obtained, the fixation point is further optimized according to the visual comfort, the subject drives according to the fixation point, the judgment and the scoring are carried out, and then the recommended safe driving scheme is finally determined. Saccade cluster maps under different circumstances as shown in figures 8 to 9. The rightmost of images 8 through 9 show the saccade cluster maps after the use of the visual safety orientation and recommendation.
Specifically, as shown in fig. 8, under the highway environment road driving task, the red circle region (7) represents a cluster of saccades of novice drivers; green circular areas (8) represent saccadic clustering within a visual safety range; the black circular area (9) represents saccade clustering after simultaneous use of visual safety orientation and recommendation strategy. As shown in fig. 9, under the urban environment road task, the red circular area (10) is clustered with saccades of novice drivers; clustering saccades for green circular regions (11) within a visual safety range; the black circular areas (12) represent saccade clusters after simultaneous use of visual safety orientation and recommendation strategy. And finally, determining a recommended eye saccade scheme by comprehensively judging and scoring the novice driver, the visual safety range and the vision safety range under different environments.
Firstly, an original driving video is collected, and a fixation point in the original driving video is positioned by adopting an attention model. Secondly, the saccade recommended by the subject is measured by a saccade instrument, the time delay between the saccade locus predicted by the model under different driving conditions and the saccade locus of the subject is analyzed, and the time delay between the saccade locus and the subject control is predicted by the model. The visual safety distance is measured by the total delay to determine a preliminary solution. And then four key features of saccades are extracted from the original driving video, the fixation point is optimized, the finally recommended safe driving scheme is determined, and reasonable recommendation can be provided for future autonomous driving.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A safe driving recommendation method in an urban environment based on a driver eye saccade is characterized by comprising the following steps:
step 1: acquiring and preprocessing an original driving video, acquiring the original driving video of a virtual and real city driving scene, and preprocessing the original driving video;
step 2: obtaining an eye-movement track and a fixation point of a subject and a model, calling the subject with driving experience to participate in an eye-movement experiment, obtaining the eye-movement track common to the subject, and obtaining the eye-movement track common to the subject through an eye-movement equipment; selecting an attention model to predict a fixation point and a corresponding glancing track of the preprocessed driving video;
and step 3: preliminarily selecting a glance trajectory scheme, and preliminarily selecting an optimal scheme of the glance trajectory by comparing the fixation points of the subject and the model with the glance differences;
and 4, step 4: extracting key features related to saccades in the preprocessed original driving video;
and 5: constructing a neural network model, selecting a most reasonable fixation point, constructing the neural network model, and selecting the most reasonable fixation point by referring to an overall safe driving scheme to meet the requirement of safe driving;
step 6: and (5) optimizing the fixation point in the step (5), determining a final recommended safe driving scheme, optimizing the fixation point selected in the step (5) in one step, improving driving comfort, and enabling the subject to participate in the scheme through subjective evaluation to finally promote the goal of safe driving.
2. The driver saccade-based safe driving recommendation method in urban environments according to claim 1, wherein the original driving video acquisition of virtual and real urban driving scenes comprises: half of the original driving video is shot under the field condition, the other half is shot under the virtual scene, the original driving video adopts a digital camera arranged on the windshield of the automobile, and in addition, a high-definition camera is used for shooting, and the frames per second are 30.
3. The driver eye glance-based recommendation method for safe driving in urban environment as claimed in claim 2, wherein said digital camera record installed on the windshield of the car extracts 6 high-definition color video clips from the driving environment of 10 expressways and urban roads in real and virtual scenes.
4. The method for recommending safe driving in urban environment based on saccades of drivers of claim 3, wherein said preprocessing of the original driving video comprises color thresholding and Gaussian filtering of the original driving video.
5. The driver saccade-based safe driving recommendation method in the urban environment according to claim 4, wherein the step 2 specifically comprises: the subjects at least comprise 35 subjects, all the subjects have driving experience for at least one year or driving records of 10000 km or more, and all the subjects have normal vision or are corrected to be normal and have normal color vision.
6. The driver saccade-based safe driving recommendation method in the urban environment according to claim 1, wherein the step 2 specifically comprises: the attention model uses a classical attention model, and 2 different saccadic eye scan schemes of the subject's gaze point and the model predicted gaze point are obtained by analyzing the pre-processed original driving video using the classical attention model.
7. The method for recommending safe driving in urban environment based on saccades of drivers according to claim 1, wherein step 3 comprises 4 key features related to saccades, said key features comprising: four characteristics of the difference of the significant markers of the two adjacent frames designated by the driving expert, the distance difference between the eye movement scanning viewpoint of the subject and the eye movement scanning viewpoint of the model between the two adjacent frames, the distance difference between the two eye movement scanning schemes of the same frame and the distance difference between the eye movement scanning viewpoint of the subject and the eye movement scanning viewpoint of the model between the forward and backward videos and the difference between the optimal significant marker and the suboptimal significant marker are analyzed.
8. The driver saccade-based safe driving recommendation method in urban environment according to claim 1, characterized in that the key feature extraction mode is as follows:
step 4.1: obtaining saccade data of the subject and the model through saccade instrument and model prediction;
step 4.2: and extracting the saccades by adopting wavelet analysis and an approximate entropy algorithm.
9. The driver saccade-based safe driving recommendation method in the urban environment according to claim 1, characterized in that the step 5 specifically comprises: and selecting a forward propagation network training, prejudging and verifying a sweep point in the residual preprocessed original driving video according to the key characteristics.
10. The driver saccade-based safe driving recommendation method in the urban environment according to claim 1, characterized in that the step 6 specifically comprises: and after the recommended point scheme in the step 5 is obtained, selecting to further optimize the fixation point according to the visual comfort and safety, driving the subject according to the recommended fixation point, and then participating in judgment and scoring.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110854240.8A CN113569733A (en) | 2021-07-28 | 2021-07-28 | Safe driving recommendation method based on saccade of driver in urban environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110854240.8A CN113569733A (en) | 2021-07-28 | 2021-07-28 | Safe driving recommendation method based on saccade of driver in urban environment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113569733A true CN113569733A (en) | 2021-10-29 |
Family
ID=78168192
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110854240.8A Pending CN113569733A (en) | 2021-07-28 | 2021-07-28 | Safe driving recommendation method based on saccade of driver in urban environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113569733A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114333489A (en) * | 2021-12-30 | 2022-04-12 | 广州小鹏汽车科技有限公司 | Remote driving simulation method, device and simulation system |
-
2021
- 2021-07-28 CN CN202110854240.8A patent/CN113569733A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114333489A (en) * | 2021-12-30 | 2022-04-12 | 广州小鹏汽车科技有限公司 | Remote driving simulation method, device and simulation system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110487562B (en) | Driveway keeping capacity detection system and method for unmanned driving | |
| JP7332726B2 (en) | Detecting Driver Attention Using Heatmaps | |
| US9384401B2 (en) | Method for fog detection | |
| Kühnl et al. | Monocular road segmentation using slow feature analysis | |
| CN110222667B (en) | Open road traffic participant data acquisition method based on computer vision | |
| CN103824037B (en) | Vehicle anti-tracking alarm device | |
| DE102009048699A1 (en) | Travel's clear path detection method for motor vehicle i.e. car, involves monitoring images, each comprising set of pixels, utilizing texture-less processing scheme to analyze images, and determining clear path based on clear surface | |
| CN108446678A (en) | A kind of dangerous driving behavior recognition methods based on skeleton character | |
| DE102009050492A1 (en) | Travel's clear path detection method for motor vehicle i.e. car, involves monitoring images, each comprising set of pixels, utilizing texture-less processing scheme to analyze images, and determining clear path based on clear surface | |
| DE102009048892A1 (en) | Clear traveling path detecting method for vehicle e.g. car, involves generating three-dimensional map of features in view based upon preferential set of matched pairs, and determining clear traveling path based upon features | |
| CN105835880A (en) | Lane tracking system | |
| CN108447303A (en) | The periphery visual field dangerous discernment method coupled with machine vision based on human eye vision | |
| CN111047874B (en) | Intelligent traffic violation management method and related product | |
| CN112016429A (en) | Fatigue driving detection method based on train cab scene | |
| CN109063667A (en) | A kind of video identification method optimizing and method for pushing based on scene | |
| CN109191492B (en) | Intelligent video black smoke vehicle detection method based on contour analysis | |
| CN114913449A (en) | Road surface camera shooting analysis method for Internet of vehicles | |
| CN103481842B (en) | A kind of changing method of moving vehicles detection and tracking pattern | |
| CN111626221A (en) | Driver gazing area estimation method based on human eye information enhancement | |
| CN105989614A (en) | Dangerous object detection method based on fusion of multi-source visual information | |
| CN110516602A (en) | A kind of public traffice passenger flow statistical method based on monocular camera and depth learning technology | |
| CN113569733A (en) | Safe driving recommendation method based on saccade of driver in urban environment | |
| CN116012822B (en) | Fatigue driving identification method and device and electronic equipment | |
| JP2021130390A (en) | Driver state estimation device | |
| JP2021130389A (en) | Driver state estimation device |
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 |