[go: up one dir, main page]

WO2023132800A2 - Procédé d'anticipation, de détection et de prévention dans un environnement 3d - Google Patents

Procédé d'anticipation, de détection et de prévention dans un environnement 3d Download PDF

Info

Publication number
WO2023132800A2
WO2023132800A2 PCT/TR2022/050011 TR2022050011W WO2023132800A2 WO 2023132800 A2 WO2023132800 A2 WO 2023132800A2 TR 2022050011 W TR2022050011 W TR 2022050011W WO 2023132800 A2 WO2023132800 A2 WO 2023132800A2
Authority
WO
WIPO (PCT)
Prior art keywords
route
vehicle
camera images
machine learning
camera
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.)
Ceased
Application number
PCT/TR2022/050011
Other languages
English (en)
Other versions
WO2023132800A3 (fr
Inventor
Nazli TEMUR
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xena Vision Yazilim Ve Savunma AS
Original Assignee
Xena Vision Yazilim Ve Savunma AS
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xena Vision Yazilim Ve Savunma AS filed Critical Xena Vision Yazilim Ve Savunma AS
Priority to PCT/TR2022/050011 priority Critical patent/WO2023132800A2/fr
Publication of WO2023132800A2 publication Critical patent/WO2023132800A2/fr
Publication of WO2023132800A3 publication Critical patent/WO2023132800A3/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/04—Detecting movement of traffic to be counted or controlled using optical or ultrasonic detectors
    • 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
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G08G1/096708—Systems involving transmission of highway information, e.g. weather, speed limits where the received information might be used to generate an automatic action on the vehicle control
    • G08G1/096716—Systems involving transmission of highway information, e.g. weather, speed limits where the received information might be used to generate an automatic action on the vehicle control where the received information does not generate an automatic action on the vehicle control
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G08G1/096766—Systems involving transmission of highway information, e.g. weather, speed limits where the system is characterised by the origin of the information transmission
    • G08G1/096783—Systems involving transmission of highway information, e.g. weather, speed limits where the system is characterised by the origin of the information transmission where the origin of the information is a roadside individual element
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G08G1/096766—Systems involving transmission of highway information, e.g. weather, speed limits where the system is characterised by the origin of the information transmission
    • G08G1/096791—Systems involving transmission of highway information, e.g. weather, speed limits where the system is characterised by the origin of the information transmission where the origin of the information is another vehicle
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/16—Anti-collision systems
    • G08G1/161—Decentralised systems, e.g. inter-vehicle communication
    • G08G1/162—Decentralised systems, e.g. inter-vehicle communication event-triggered
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/16—Anti-collision systems
    • G08G1/164—Centralised systems, e.g. external to vehicles
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/16—Anti-collision systems
    • G08G1/166—Anti-collision systems for active traffic, e.g. moving vehicles, pedestrians, bikes
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G5/00—Traffic control systems for aircraft
    • G08G5/20—Arrangements for acquiring, generating, sharing or displaying traffic information
    • G08G5/22—Arrangements for acquiring, generating, sharing or displaying traffic information located on the ground
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G5/00—Traffic control systems for aircraft
    • G08G5/20—Arrangements for acquiring, generating, sharing or displaying traffic information
    • G08G5/26—Transmission of traffic-related information between aircraft and ground stations
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G5/00—Traffic control systems for aircraft
    • G08G5/80—Anti-collision systems
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/095—Traffic lights

Definitions

  • the invention disclosed is a method for recognizing routes, generating a projection for the recognized route and generating warnings from the said projections based on visual camera data.
  • the method comprises a general understanding of route in three dimensional environments and can be employed for air, sea, land or a selection of the said.
  • An example of the said approach can be monitoring vehicle parameters for accident detection. For example, if a vehicle entered a pre-defined zone over a prescribed speed limit and committed an accident, these type of methods label all vehicles over the very same speed limit a warning. But in today’s world we are sure about a probabilistic approach; increased speed only increases the likelihood of an accident, does not stipulate one.
  • the invention disclosed is a method for accident anticipation and prevention in a three dimensional space.
  • the developments in the technology brings accident anticipation in three dimensions a must, simply because vehicles using land, air and sea exist more often. Additionally, vehicles use close proximity and low altitude in the air and land and take off more frequently. This is basically due to developments in hybrid vehicles, drones and flying cars. Besides, a traction in three dimensions gives more accurate results in each and every area, even if the motion is to be assumed as two dimensional.
  • the method disclosed tracks a vehicle on at least one camera (100).
  • the camera mentioned herein can be a on-board camera, a surveillance camera, a drone camera, a traffic inspection camera that is stationary or on the move.
  • This camera data is converted to vehicle tracking data and a path projection is generated based on the path detected via machine learning.
  • the method generates a warning and extracts this warning to user display in any form to prevent a collision.
  • the invention trains a machine learning method based on variance of the route of selected vehicles. These vehicles are classified on their movement trajectories and these trajectories are inputted into a learning model.
  • the said model does not target anticipation, but a trajectory of the vehicle.
  • This training is based on variance of displacement in three dimensions and aims to generate a model that correlates with the real displacement the target vehicle does.
  • the invention described Having trained this model on real world data achieving a correlation, the invention described generates a path based on classification of vehicle movement. It classifies vehicle movement with similar variances on speed, maneuvering and the like. This is done at the site, preferably on the camera and prevents a video transmission.
  • the invention disclosed Having trained a model for variance and correlation of displacement, the invention disclosed generates at least one route projection for each incoming vehicle.
  • This projection can either be transferred into a remote server or can be processed on site.
  • Transforming a remote server can be an option, possibly aimed at processing at least one data obtained from camera.
  • the said accident is expected to be against surroundings; the vehicle can hit objects at surroundings, get out of the safe route, roll over, sink, fall or the like. Any route based anomaly that appears on variance and correlation of the route would count as anomaly, thus accident in this case.
  • the invention disclosed monitors and classifies the accident analytics via training a machine learning model based on safe actions of previous routes. To do so, the invention disclosed feeds a camera input to the learning algorithm to generate a set of limits for safe route.
  • This safe route is on three dimensions, thus can be used for land, air, sea vehicles or a selection from these. Any route out of these limits are marked at points of intersection after a curve is formed out of them.
  • the invention disclosed transforms images into paths and generates a path projections based on the path taken.
  • the invention disclosed generates a path projection and handles accident anticipation based on this projection; thus capable of processing the video stream either on the camera, on a remote location or a selection from these.
  • the path generation and projection is unique as described within the invention disclosed, the latter may vary within application as preferred by a person skilled in the art.
  • the machine learning referred in this invention is employed accordingly; a camera data is retrieved including multiple passing of the path to be observed.
  • the camera data is extracted to path data and a model is trained onto this data with regards to selected parameters such as speed (linear and angular), position, angles or rotation; yaw, roll, pitch; acceleration or a selection from the said variables.
  • the said warning is a sign that will alert the person in command of the vehicle that he is under strong probability of an incident. This can be done in many ways; a sign can be utilized both inside and outside of the vehicle, a digital warning aimed at a digital device can be issued and other known forms of warning a driver/pilot can be utilized. The characteristics of warning can be easily modified by a person skilled in the art. Besides, more than one method of warning can be utilized simultaneously.
  • the first advantageous effect of the invention is its capability of foreseeing any anticipation or anomaly before it occurs; with a high accuracy and precision in well time. This leads to customer warnings to prevent an accident including but not limited to speed reduction, maneuver variation or any other type of anticipation prevention.
  • the next advantageous effect of the invention is rapid transmission of the information as retrieved from the camera image. Especially in the data fusion case, transmission of the variance of the route and the trajectory extracted from the said variance is transferred, so that a timely warning can be issued.
  • Yet another advantageous effect of the invention is GPS correction.
  • the invention disclosed can easily correct the deficiencies or inaccuracies in the GPS signal. For instance, the invention disclosed can easily generate a position indicator along the GPS for times of lost signal based on the trajectory analyzed. This indication of location can be transferred as a regular location data, irrespective of the received signal quality or accuracy.
  • the other advantageous effect of the invention disclosed is its capability of three dimensional working.
  • new types of vehicles take presence and have multiple issues on security, such as flying cars.
  • aerial transport forms increase, their need of security becomes more and more important as accident are vital.
  • the invention disclosed proposes a accident anticipation system for any type of vehicle; from bikes to cars; from drones to flying cars; planes, helicopters, water vehicles and even pedestrians and scooters.
  • Any moving object capable of carrying a camera can use the invention disclosed without additional constraints, as the machine learning model will be trained on the vehicle itself but not an arbitrary function (i.e. median) of other vehicles present.
  • the use of machine learning is thus not an arbitrary preference, but a compliance for any type of vehicle that can anticipate with others.
  • FIG. 1 is a schematic of the invention disclosed as a working model. The figure is explain the working model of the invention disclosed.
  • the first embodiment of the invention is its use with single camera, being located on the vehicle and having a processor unit at the same location.
  • the processing unit retrieves the image from the camera instantly and transforms the image to trajectories in three dimensions via a machine leaning algorithm. To do so, it uses a pre-trained model of variances of route parameters, such as speed, route, maneuvering, angular velocity and the like to make a projection of future route. From these parameters of the future route, the invention disclosed makes a comparison and detects anomalies, if any. In the case of anomalies in the projected route, it generates an alert to the driver. Besides, this
  • the second embodiment of the invention disclosed comprises its use on a single camera being located outside of the vehicle and having a processor unit on the said camera.
  • the image taken by the camera is processed to extract the variance of the route as explained and generates a expected route for the tracked vehicle.
  • This route can be transferred to another location, or can be merged with other expected routes within the same processing unit, so that a warning can be issued at an expected anticipation.
  • the third embodiment of the invention disclosed comprises a single camera, a processing unit onboard and a GPS unit connected to the processing unit.
  • images are retrieved from the camera and sent to the processing unit and a location correction for the incoming GPS data is made, in accordance with the path data as processed from the image.
  • location data from the camera image takes prevalence, as an accurate path recognition has been made.
  • route variances are calculated and a route projection is generated as usual.
  • a warning is generated and transmitted to prevent an incident.
  • the term GPS is used to cover all location data source systems including satellite and stationary location sensing systems, equipment or methods.
  • the invention disclosed comprises plurality of cameras, all located on the vehicle for distinct angles of view and processing unit on the vehicle.
  • multiple path recognition is done based on multiple streams from camera to obtain a single path and variance of the calculated path.
  • An estimation based on these camera images are conducted on the said path and variances calculated.
  • the invention disclosed comprises plurality of cameras, outside of the vehicle.
  • cameras mentioned can be either stationary (i.e. surveillance cameras) or mobile (on the car or drone cameras). Regardless of the camera type and location, there are two possibilities for this embodiment.
  • the camera images are either processed on the camera location or transferred directly to a remote location for processing.
  • calculated routes are sent to a central computer for analysis and a path projection is made via calculating variances of the route(s).
  • the sixth embodiment of the invention comprises at least one camera at the vehicle and one camera outside the vehicle.
  • the camera(s) outside the vehicle can be stationary or fixed. While aiming at the vehicle tube tracked, the all cameras generate a path data for the vehicle.
  • Each camera transmits this path data to a control computer at a desired location, preferably within the vehicle. Else, this location can be outside the vehicle; a computer on a road sign or control tower can be easily applied by a person skilled in the art.
  • the control computer calculates path variances and utilizes the predictive model for the route at hand and generates a projection. When projection of route intercepts shows an anomaly, the control computer triggers a warning to the person in command of the vehicle, to prevent an incident.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Atmospheric Sciences (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Traffic Control Systems (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)
  • Geophysics And Detection Of Objects (AREA)

Abstract

L'invention concerne un procédé de reconnaissance d'itinéraires, de génération d'une projection de l'itinéraire reconnu et de génération d'avertissements à partir desdites projections sur la base de données visuelles de caméra. Le procédé comprend une compréhension générale de l'itinéraire dans des environnements tridimensionnels et s'emploie dans les domaines aériens et/ou maritimes et/ou terrestres.
PCT/TR2022/050011 2022-01-06 2022-01-06 Procédé d'anticipation, de détection et de prévention dans un environnement 3d Ceased WO2023132800A2 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/TR2022/050011 WO2023132800A2 (fr) 2022-01-06 2022-01-06 Procédé d'anticipation, de détection et de prévention dans un environnement 3d

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/TR2022/050011 WO2023132800A2 (fr) 2022-01-06 2022-01-06 Procédé d'anticipation, de détection et de prévention dans un environnement 3d

Publications (2)

Publication Number Publication Date
WO2023132800A2 true WO2023132800A2 (fr) 2023-07-13
WO2023132800A3 WO2023132800A3 (fr) 2023-08-17

Family

ID=87074377

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/TR2022/050011 Ceased WO2023132800A2 (fr) 2022-01-06 2022-01-06 Procédé d'anticipation, de détection et de prévention dans un environnement 3d

Country Status (1)

Country Link
WO (1) WO2023132800A2 (fr)

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108154681B (zh) * 2016-12-06 2020-11-20 杭州海康威视数字技术股份有限公司 发生交通事故的风险预测方法、装置及系统
KR102030583B1 (ko) * 2017-11-23 2019-10-11 (주)에이텍티앤 인공지능 기반의 교통사고 예측시스템 및 그 방법
US20190354838A1 (en) * 2018-05-21 2019-11-21 Uber Technologies, Inc. Automobile Accident Detection Using Machine Learned Model

Also Published As

Publication number Publication date
WO2023132800A3 (fr) 2023-08-17

Similar Documents

Publication Publication Date Title
Djuric et al. Uncertainty-aware short-term motion prediction of traffic actors for autonomous driving
Biparva et al. Video action recognition for lane-change classification and prediction of surrounding vehicles
CN112351928B (zh) 基于轨道分段的铁路障碍物检测的方法与系统
EP3807806B1 (fr) Système et procédé de détermination d'une intention d'un objet par l'intermédiaire d'attributs visuels
US11465652B2 (en) Systems and methods for disengagement prediction and triage assistant
US11927967B2 (en) Using machine learning models for generating human-like trajectories
US10496099B2 (en) Systems and methods for speed limit context awareness
CN110920609B (zh) 用于模仿前车的系统和方法
US11003928B2 (en) Using captured video data to identify active turn signals on a vehicle
US20220188695A1 (en) Autonomous vehicle system for intelligent on-board selection of data for training a remote machine learning model
CN115038628A (zh) 对象速度和/或偏航率检测和跟踪
CN117813227A (zh) 用于自主交通工具的互补控制系统
EP4170450A1 (fr) Procédé et système de commutation entre instructions de guidage local et a distance pour un véhicule autonome
CN116745188A (zh) 基于不确定道路使用者的行为生成自主车辆的纵向计划的方法和系统
US20250304109A1 (en) Systems and methods of determining changes in pose of an autonomous vehicle
US20250078519A1 (en) Road geometry estimation for vehicles
US12221115B1 (en) Self-supervised global velocity determination for perception system
Virdi Using deep learning to predict obstacle trajectories for collision avoidance in autonomous vehicles
WO2022108744A1 (fr) Système de rétroaction embarqué pour véhicules autonomes
WO2023132800A2 (fr) Procédé d'anticipation, de détection et de prévention dans un environnement 3d
Rovid et al. Digital twin and cloud based remote control of vehicles
Deore et al. A new approach for navigation and traffic signs indication using map integrated augmented reality for self-driving cars
Siddiqui et al. Object/Obstacles detection system for self-driving cars
Ashraf et al. V-CAS: A Realtime Vehicle Anti Collision System Using Vision Transformer on Multi-Camera Streams
Guo et al. Toward human-like lane following behavior in urban environment with a learning-based behavior-induction potential map

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 07/10/2024)

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22919128

Country of ref document: EP

Kind code of ref document: A2

122 Ep: pct application non-entry in european phase

Ref document number: 22919128

Country of ref document: EP

Kind code of ref document: A2