EP4584755A1 - Video-based automated driving conditions classification system and method - Google Patents
Video-based automated driving conditions classification system and methodInfo
- Publication number
- EP4584755A1 EP4584755A1 EP22812766.8A EP22812766A EP4584755A1 EP 4584755 A1 EP4584755 A1 EP 4584755A1 EP 22812766 A EP22812766 A EP 22812766A EP 4584755 A1 EP4584755 A1 EP 4584755A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- module
- vehicle
- data set
- imagery
- resulting
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
Definitions
- the third resulting data set comprises drive conditions characteri zation of static and dynamic elements of the surroundings of the vehicle , comprised of at least one of an annotated frames and/or video and/or critical event and/or plots and/or full test drive report , etc .
- the present invention also describes a system for imagery classification of the surrounding driving conditions of a vehicle according to previous described method, comprising a vision module, connected to an AT module which in turn is connected to an analysis module; the vision module is data feed by at least one image acquisition camera installed in at least one structure and surrounding position of the vehicle; characterized by enabling the forwarding of a resulting data set to a module, comprising drive conditions characterization of static and dynamic elements of the surroundings of the vehicle.
- Performance metrics such as the positioning error in dynamic conditions, are difficult to evaluate, being necessary a high-grade reference system based on the combination of, e.g., multiple sensors, post-processing techniques, Global Navigation Satellite System (GNSS) correction services, to obtain high accuracy ground-truth in an order of magnitude better than the System Under Evaluation (SUE) .
- GNSS Global Navigation Satellite System
- SUE System Under Evaluation
- the SUE will be affected by multiple surrounding conditions and events, that can have severe impact in the overall performance.
- a well-known example is the effect of surrounding buildings (e.g., urban canyons) that block and obstruct GNSS signals, leading to a severe impact in the accuracy and availability. Therefore, the performance obtained in a test drive will depend on the static and/or dynamic surrounding elements and conditions.
- the above-mentioned evaluation setup allows to measure performance metrics like positioning or heading error, it does not provide the context to the test drive conditions. This might lead to a poor evaluation because the SUE can be evaluated mostly in favorable and optimistic conditions, and the performance reported may not hold in more complex or demanding scenarios. Detailed information about the conditions or events encountered during the evaluation test drive is essential to give context to the performance metrics.
- mapping considers the use of several types of sensor data, to detect and track dynamic objects.
- the dynamic objects include, for example, other vehicles, pedestrians, bicycles, scooters, and so on.
- sensors e.g., LiDAR, different types of cameras, radar, etc
- current technological developments lack relation with test drive conditions characterization and automated tagging, being, therefore, only focused on the detection, tracking and mapping of dynamic objects, lacking, therefore, the detection of static elements of the surrounding environment. Once the main focus resorts on the mapping, the analysis of the surrounding conditions, e.g., conditions density statistics, remain unattended.
- the present invention describes a new solution for automatic classification / tagging of vehicle test drive conditions.
- the proposed solution uses video feed(s) from one or more cameras mounted in a vehicle and explores image processing, computer vision, machine learning and analysis techniques to automatically classify the static and dynamic conditions surrounding the vehicle during a test drive, therefore leading to a higher quality evaluation process.
- Additional sensors can also be used to obtain more information, e.g., Global Navigation Satellite System (GNSS) , Inertial Navigation System (INS) , Light Detection and Ranging (LiDAR) , Radar or Speed.
- GNSS Global Navigation Satellite System
- INS Inertial Navigation System
- LiDAR Light Detection and Ranging
- Radar or Speed e.g., Radar or Speed.
- the resulting obtained information allows to automatically determine data about the percentage of time / performance in each analysed scenario which is being affected by a specific condition (e.g., buildings / urban canyons, vegetation and open sky) or dynamic obstacles in the surroundings of the vehicle (such as detection of trucks, cars and buses) .
- the resulting data also allows establishing the evaluation test drive profile, providing a context overview to the performance metrics, allowing to rapidly verify the test drive conditions.
- This herein disclosed solution can be used to determine and provide context in evaluation processes with the identification of sources of performance degradation during long test drives, and can also be useful in other applications, such as in the research and development process to identify critical events where the SUE underperforms, i.e., the APS system.
- the proposed solution is configured to automatically extract detailed context data for an entire test drive, and provide information such as:
- Instantaneous information for the surrounding conditions of the vehicle at each moment e.g., one camera for multiple sectors of the vehicle, or a 360 degrees camera
- Type of scenario where the vehicle is currently located e.g., urban canyon, highway, tunnel, etc.
- the proposed solution aims to solve a problem that is identi fied for quite some time in the area of test driving for highly automated driving technology, which can be integrated into high-quality global reference systems like high-grade GNSS+ INS to perform rigorous characterisation of automotive positioning systems , without any additional ef fort .
- the solution structure can be divided into three main modules , a vision module , an Al module and an analysis module .
- the vision module is responsible for the video and image capturing and processing operations , pre and post processing .
- the Al module is responsible for the scene classi fication exploring two types of neural networks .
- the analysis module receives the Al module output and is responsible for the extracting relevant raw information, context information and all statistical analysis.
- This analysis module is also responsible for providing the obtained data in different formats (e.g., Comma Separated Values (CSV) , charts, annotated f rames/video, full test drive statistical reports, etc) .
- This data comprises the test conditions identified by the solution during the respective test drive and is used in the generation of the route profile report. This report allows comparison between performance results from different SUEs in different test drives and environments.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Traffic Control Systems (AREA)
Abstract
The present application describes a video-based automated driving conditions classification system and method. The proposed method for imagery classification of the surrounding driving conditions of a vehicle comprises executing imagery processing on a vision module (201) which is data feed by at least one image acquisition camera (10) installed in at least one structure and surrounding position of the vehicle; collecting and processing a first resulting imagery data set on an Al module (202) which is data feed by the vision module (201); extracting and processing a second resulting imagery data set on an analysis module (203) which is data feed by the Al module (202); exporting a third resulting data set to a results module (30) which comprises drive conditions characterization of static elements of the surroundings of the vehicle.
Description
DESCRIPTION "VIDEO-BASED AUTOMATED DRIVING CONDITIONS CLASSIFICATION SYSTEM AND METHOD"
Technical Field
The present application describes a video-based automated driving conditions classi fication system and method .
Background art
Patent document US 10753750B2 describes a system, methods , and other embodiments related to improving mapping of a surrounding environment by a mapping vehicle . In one embodiment , a method includes identi fying dynamic obj ects within the surrounding environment that are proximate to the mapping vehicle from sensor data of at least one sensor o f the mapping vehicle . The dynamic obj ects are trackable obj ects that are moving within the surrounding environment . The method includes generating paths of the dynamic obj ects through the surrounding environment relative to the mapping vehicle according to separate observations of the dynamic obj ects embodied within the sensor data . The method includes producing a map of the surrounding environment from the paths .
Summary
The present invention describes a method for imagery classi fication of the surrounding driving conditions of a vehicle comprising executing imagery processing on a vision module which is data feed by at least one image acquisition camera installed in at least one structure and surrounding
position of the vehicle ; collecting and processing a first resulting imagery data set on an Al module which is data feed by the vision module ; extracting and processing a second resulting imagery data set on an analysis module which is data feed by the Al module ; exporting a third resulting data set to a results module which comprises drive conditions characteri zation of static and dynamic elements of the surroundings of the vehicle .
In a proposed embodiment of present invention, the imagery processing comprises at least one of a conversion, formatting, downscaling, upscaling or filtering .
Yet in another proposed embodiment of present invention, the first resulting imagery data set is adapted to match, adapt , and improve the Al module based on the at least one image acquisition camera video data characteristics like resolution, frame rate or image format .
Yet in another proposed embodiment of present invention, the second resulting imagery data set comprises at least one relevant element or condition of the surroundings of the vehicle .
Yet in another proposed embodiment of present invention, the Al module comprises at least two neuronal networks , a Convolutional Neuronal Network and a Deep Neuronal Network .
Yet in another proposed embodiment of present invention, the Deep Neuronal Network is configured to identi fy multiple classes and/or perform semantic segmentation .
Yet in another proposed embodiment of present invention, the multiple classes represent at least one relevant element or condition of the surroundings of the vehicle including at least one of a sky view and/or buildings and/or vegetation and/or tunnels and/or bridges and/or underground parks and/or highway traf fic signs , etc .
Yet in another proposed embodiment of present invention, the semantic segmentation comprises identi fying pixels of the first resulting imagery data set that belong to the multiple classes .
Yet in another proposed embodiment of present invention, the Convolutional Neuronal Network is configured to identi fy and/or count multiple static and/or dynamic obj ects including at least one of a nearby vehicle and/or traf fic sign, etc .
Yet in another proposed embodiment of present invention, the extraction and proces sing of the second resulting imagery data set comprises at least one of a data cross-checking and/or outlier removal and/or class density calculation and/or analysis per side and/or analysis per camera and/or aggregated analysis and/or scenario type inference and/or critical event flags and/or test drive condition statistics generation .
Yet in another proposed embodiment of present invention, the third resulting data set comprises drive conditions characteri zation of static and dynamic elements of the surroundings of the vehicle , comprised of at least one of an annotated frames and/or video and/or critical event and/or plots and/or full test drive report , etc .
The present invention also describes a system for imagery classification of the surrounding driving conditions of a vehicle according to previous described method, comprising a vision module, connected to an AT module which in turn is connected to an analysis module; the vision module is data feed by at least one image acquisition camera installed in at least one structure and surrounding position of the vehicle; characterized by enabling the forwarding of a resulting data set to a module, comprising drive conditions characterization of static and dynamic elements of the surroundings of the vehicle.
General Description
Automotive Positioning Systems (APS) aim to provide reliable and high-performance results, e.g., within the decimeter to centimeter level accuracy, in complex and dynamic real-world conditions in order to support Highly Automated Driving (HAD) or Autonomous Driving (AD) . Therefore, recurring evaluation in long test drives needs to be performed to ensure that the performance and safety requirements are achieved in diverse scenarios, and not only for best-case scenarios.
Performance metrics, such as the positioning error in dynamic conditions, are difficult to evaluate, being necessary a high-grade reference system based on the combination of, e.g., multiple sensors, post-processing techniques, Global Navigation Satellite System (GNSS) correction services, to obtain high accuracy ground-truth in an order of magnitude better than the System Under Evaluation (SUE) .
During test drive procedures in different scenarios and environments, the SUE will be affected by multiple surrounding conditions and events, that can have severe impact in the overall performance. A well-known example is the effect of surrounding buildings (e.g., urban canyons) that block and obstruct GNSS signals, leading to a severe impact in the accuracy and availability. Therefore, the performance obtained in a test drive will depend on the static and/or dynamic surrounding elements and conditions. Therefore, although the above-mentioned evaluation setup allows to measure performance metrics like positioning or heading error, it does not provide the context to the test drive conditions. This might lead to a poor evaluation because the SUE can be evaluated mostly in favorable and optimistic conditions, and the performance reported may not hold in more complex or demanding scenarios. Detailed information about the conditions or events encountered during the evaluation test drive is essential to give context to the performance metrics.
As previously disclosed, existing prior art systems and methods for a vehicle mapping through inferences of observed objects only suggest improving mapping of a surrounding environment by a mapping vehicle. These mapping considers the use of several types of sensor data, to detect and track dynamic objects. The dynamic objects include, for example, other vehicles, pedestrians, bicycles, scooters, and so on. Although it is suggested the use of several types of sensors (e.g., LiDAR, different types of cameras, radar, etc) , current technological developments lack relation with test drive conditions characterization and automated tagging, being, therefore, only focused on the detection, tracking and mapping of dynamic objects, lacking, therefore, the
detection of static elements of the surrounding environment. Once the main focus resorts on the mapping, the analysis of the surrounding conditions, e.g., conditions density statistics, remain unattended.
Methods to complement performance metrics with detailed context information about the different conditions or scenarios encountered during the evaluation test drive are therefore of utmost importance. The context regarding the test drive conditions is usually obtained manually or using map-matching techniques. Taking into consideration the long test drives that are required to evaluate an APS, manually obtaining this type of detailed information is unfeasible. In addition, the level of context detail obtained with this approach is very low. The map-matching approach allows automating the process but also provides a low level of detail .
Considering the above, the present invention describes a new solution for automatic classification / tagging of vehicle test drive conditions. The proposed solution uses video feed(s) from one or more cameras mounted in a vehicle and explores image processing, computer vision, machine learning and analysis techniques to automatically classify the static and dynamic conditions surrounding the vehicle during a test drive, therefore leading to a higher quality evaluation process. Additional sensors can also be used to obtain more information, e.g., Global Navigation Satellite System (GNSS) , Inertial Navigation System (INS) , Light Detection and Ranging (LiDAR) , Radar or Speed.
The resulting obtained information allows to automatically determine data about the percentage of time / performance in
each analysed scenario which is being affected by a specific condition (e.g., buildings / urban canyons, vegetation and open sky) or dynamic obstacles in the surroundings of the vehicle (such as detection of trucks, cars and buses) . The resulting data also allows establishing the evaluation test drive profile, providing a context overview to the performance metrics, allowing to rapidly verify the test drive conditions.
This herein disclosed solution can be used to determine and provide context in evaluation processes with the identification of sources of performance degradation during long test drives, and can also be useful in other applications, such as in the research and development process to identify critical events where the SUE underperforms, i.e., the APS system.
The proposed solution is configured to automatically extract detailed context data for an entire test drive, and provide information such as:
Instantaneous information for the surrounding conditions of the vehicle at each moment (e.g., one camera for multiple sectors of the vehicle, or a 360 degrees camera) ;
Type of scenario where the vehicle is currently located, (e.g., urban canyon, highway, tunnel, etc.)
- Aggregated information in an overview route profile, e.g., in a percentage per scenario statistics summary;
- Detailed timeline analysis allowing to understand the evolution of each condition and identify critical events where the SUE, i.e. the APS, underperforms;
- Data exported in multiple and adaptable formats to be used in external tools to perform further analysis or processing .
The proposed solution aims to solve a problem that is identi fied for quite some time in the area of test driving for highly automated driving technology, which can be integrated into high-quality global reference systems like high-grade GNSS+ INS to perform rigorous characterisation of automotive positioning systems , without any additional ef fort .
It renders potential to contribute to setting up a high- quality service for evaluation and veri fication of designs for dynamic positioning systems in di f ferent areas where moving vehicles are used ( e . g . , automotive industry, farming vehicles , landscaping products , near coast vessels , ... ) . This service will be possible by the usage of high accuracy reference systems ( state of the art measurement systems together with post-processing) . The usage of the Video-Based Automated Driving Conditions Class i fication solution can be applied to automatically obtain detailed information about the test drive conditions . This information can be used to give context to the performance metrics , providing such service with higher quality and reduced human ef forts .
The solution structure can be divided into three main modules , a vision module , an Al module and an analysis module . The vision module is responsible for the video and image capturing and processing operations , pre and post processing . The Al module is responsible for the scene classi fication exploring two types of neural networks . The analysis module receives the Al module output and is
responsible for the extracting relevant raw information, context information and all statistical analysis. This analysis module is also responsible for providing the obtained data in different formats (e.g., Comma Separated Values (CSV) , charts, annotated f rames/video, full test drive statistical reports, etc) . This data comprises the test conditions identified by the solution during the respective test drive and is used in the generation of the route profile report. This report allows comparison between performance results from different SUEs in different test drives and environments.
Brief description of the drawings
For better understanding of the present application, figures representing preferred embodiments are herein attached which, however, are not intended to limit the technique disclosed herein.
Figure 1 - discloses a proposed embodiment of the overall block diagram of the video-based automated driving conditions classification system.
Figure 2 - discloses an illustration representing the application of the present invention in a common daily basis application, for example, in a city. The reference numbers refer to:
10 - image acquisition camera;
20 - video-based automated driving conditions classification system;
301 - buildings;
302 - vehicles (e.g. trucks) ;
303 - vegetation.
Description of Embodiments
With reference to the figures, some embodiments are now described in more detail, which are however not intended to limit the scope of the present application.
A particular embodiment of the video-based automated driving conditions classification system (20) herein disclosed comprises at least three main modules, a vision module (201) , an AT module (202) and an analysis module (203) .
The vision module (201) , in one of the preferred embodiments is data feed by at least one image acquisition camera (10) and is configured to execute image processing tasks. These tasks include but are not limited to: image conversion, formatting, down/upscaling, filtering and improvement (e.g. denoise, sharpening, etc) as required to match, adapt, improve the AT module input and considering the camera (s) used and video feed(s) characteristics, e.g. resolution, frame rate, image format. The image acquisition camera (10) as previously suggested is configured to capture single images, sets of images or full-motion video of the surroundings of the vehicle here inserted. The vehicle, in one of the proposed embodiments of the present invention can be one of a automated driving vehicle or a test driving vehicle .
The AT module (202) is comprised of two neuronal networks, a convolutional neuronal network (CNN) (2021) and a Deep
Neuronal Network (DNN) (2022) , both of said neuronal networks
being data feed by the results provided by the vision module (201) . The DNN (2022) is trained to identify multiple classes. A class represents a relevant element or condition in the applicational context, including but not limited to: sky view, buildings, vegetation (inc. trees) , tunnels, bridges, underground parks, highway traffic signs. The DNN (2022) is also configured to perform semantic segmentation, where each pixel of the image is classified as belonging to one of the classes that the DNN (2022) was trained to classify. The pixel is not classified if it is determined to not belong to one of the trained classes. In the disclosed embodiment, the DNN (2022) is the preferable method to extract information about the environment elements and conditions surrounding the vehicle. The CNN (2021) is used to perform object detection, therefore, is trained to detect, identify, and count multiple static or dynamic objects (inc. nearby vehicles, traffic signs, etc) . The CNN (2021) is the preferable method to identif y/count individual objects of the same type (e.g., trucks nearby) .
The Al module (202) will data feed the analysis module (203) which in turn will be configured to extract and combine the information provided by the two neuronal networks (2021, 2022) about relevant surrounding conditions of the vehicle captured through the image acquisition cameras (10) . The analysis module (203) operations include, but not limited to: data cross-checking, outlier removal, class density calculation, analysis per side, analysis per camera, aggregated analysis, scenario type inference, critical event flags, and test drive condition statistics generation.
The analysis module output (203) is provided to a results module (30) which can export the results in different format
(e.g., Comma-Separated Values (CSV) , annotated f rames/video, critical events, plots, full test drive report, etc) .
Claims
1. Method for imagery classification of the surrounding driving conditions of a vehicle comprising executing imagery processing on a vision module (201) which is data feed by at least one image acquisition camera (10) installed in at least one structure and surrounding position of the vehicle; collecting and processing a first resulting imagery data set on an Al module (202) which is data feed by the vision module (201) ; extracting and processing a second resulting imagery data set on an analysis module (203) which is data feed by the Al module (202) ; exporting a third resulting data set to a results module (30) which comprises drive conditions characterization of static and dynamic elements of the surroundings of the vehicle.
2. Method according to the previous claim, wherein the imagery processing comprises at least one of a conversion, formatting, downscaling, upscaling or filtering.
3. Method according to claim 1, wherein the first resulting imagery data set is adapted to match, adapt, and improve the Al module (202) based on the at least one image acquisition camera (10) video data characteristics like resolution, frame rate or image format.
4. Method according to claim 1, wherein the second resulting imagery data set comprises at least one relevant element or condition of the surroundings of the vehicle.
5. Method according to claim 1 and 3, wherein the Al module (202) comprises at least two neuronal networks, a Convolutional Neuronal Network (2021) and a Deep Neuronal Network (2022) .
6. Method according to claim 5, wherein the Deep Neuronal Network (2022) is configured to identify multiple classes and/or perform semantic segmentation.
7. Method according to claim 6, wherein the multiple classes represent at least one relevant element or condition of the surroundings of the vehicle including at least one of a sky view and/or buildings and/or vegetation and/or tunnels and/or bridges and/or underground parks and/or highway traffic signs, etc.
8. Method according to claim 6, wherein the semantic segmentation comprises identifying pixels of the first resulting imagery data set that belong to the multiple classes .
9. Method according to claim 5, wherein the Convolutional Neuronal Network (2021) is configured to identify and/or count multiple static and/or dynamic objects including at least one of a nearby vehicle and/or traffic sign, etc.
10. Method according to claim 1 and 4, wherein the extraction and processing of the second resulting imagery data set comprises at least one of a data cross-checking and/or outlier removal and/or class density calculation and/or analysis per side and/or analysis per camera and/or aggregated analysis and/or scenario type inference and/or
critical event flags and/or test drive condition statistics generation .
11. Method according to claim 1, wherein the third resulting data set comprises drive conditions characterization of static and dynamic elements of the surroundings of the vehicle, comprised of at least one of an annotated frames and/or video and/or critical event and/or plots and/or full test drive report, etc.
12. System (20) for imagery classification of the surrounding driving conditions of a vehicle according to any of the previous method claims 1 to 11, comprising a vision module (201) , connected to an Al module (202) which in turn is connected to an analysis module (203) ; the vision module (201) is data feed by at least one image acquisition camera (10) installed in at least one structure and surrounding position of the vehicle; characterized by enabling the forwarding of a resulting data set to a module (30) , comprising drive conditions characterization of static and dynamic elements of the surroundings of the vehicle.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PT11828222 | 2022-10-25 | ||
| PCT/IB2022/060468 WO2024089456A1 (en) | 2022-10-25 | 2022-10-31 | Video-based automated driving conditions classification system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4584755A1 true EP4584755A1 (en) | 2025-07-16 |
Family
ID=84361941
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP22812766.8A Pending EP4584755A1 (en) | 2022-10-25 | 2022-10-31 | Video-based automated driving conditions classification system and method |
Country Status (2)
| Country | Link |
|---|---|
| EP (1) | EP4584755A1 (en) |
| WO (1) | WO2024089456A1 (en) |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10753750B2 (en) | 2018-07-12 | 2020-08-25 | Toyota Research Institute, Inc. | System and method for mapping through inferences of observed objects |
| DE102020211280A1 (en) * | 2020-09-09 | 2022-03-10 | Zf Friedrichshafen Ag | Computer-implemented method for environment recognition for an automated driving system, machine learning method, control unit for an automated driving system and computer program for such a control unit |
-
2022
- 2022-10-31 WO PCT/IB2022/060468 patent/WO2024089456A1/en not_active Ceased
- 2022-10-31 EP EP22812766.8A patent/EP4584755A1/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| WO2024089456A1 (en) | 2024-05-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106919915B (en) | Map road marking and road quality acquisition device and method based on ADAS system | |
| US20200285913A1 (en) | Method for training and using a neural network to detect ego part position | |
| Bakirci | Advanced aerial monitoring and vehicle classification for intelligent transportation systems with YOLOv8 variants | |
| JP2022181678A (en) | Information processing system and program | |
| CN118038409B (en) | Vehicle drivable region detection method, device, electronic equipment and storage medium | |
| CN114817600A (en) | Method, device and platform for generating automatic driving simulation test scene library | |
| CN113903008A (en) | Ramp exit vehicle violation identification method based on deep learning and trajectory tracking | |
| CN114241373A (en) | End-to-end vehicle behavior detection method, system, equipment and storage medium | |
| CN118429941A (en) | Traffic signal lamp identification method, device, equipment and storage medium | |
| CN112735164B (en) | Test data construction method and test method | |
| US12553742B2 (en) | Method of labeling data obtained from vehicle sensor | |
| Certad et al. | Interaction of autonomous and manually controlled vehicles multiscenario vehicle interaction dataset | |
| KR102492290B1 (en) | Drone image analysis system based on deep learning for traffic measurement | |
| Espadaler-Clapés et al. | An accurate safety and congestion monitoring framework with a swarm of drones | |
| CN113343849A (en) | Fusion sensing equipment based on radar and video | |
| Sung et al. | Real-time traffic light recognition on mobile devices with geometry-based filtering | |
| CN112466121A (en) | Speed measuring method based on video | |
| Alam et al. | An Optimized YOLOv5 Based Approach For Real-time Vehicle Detection At Road Intersections Using Fisheye Cameras | |
| EP4584755A1 (en) | Video-based automated driving conditions classification system and method | |
| Gheorghe et al. | Road traffic analysis using unmanned aerial vehicle and image processing algorithms | |
| CN117496459A (en) | CRNet-based drone video vehicle detection method and system | |
| Mishra et al. | IDD-CRS: A Comprehensive Video Dataset for Critical Road Scenarios in Unstructured Environments | |
| Simoncini et al. | Two-stream neural architecture for unsafe maneuvers classification from dashcam videos and GPS/IMU sensors | |
| Krajewski et al. | Drone-based generation of sensor reference and training data for highly automated vehicles | |
| Rahman et al. | Automatic lane assistance line detection using yolox in vehicle speed detection system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20250411 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC ME MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
| DAV | Request for validation of the european patent (deleted) | ||
| DAX | Request for extension of the european patent (deleted) |