WO2020076264A2 - Procédé de détection de violation de ceinture de sécurité à l'aide d'images de système de reconnaissance de plaque d'immatriculation - Google Patents

Procédé de détection de violation de ceinture de sécurité à l'aide d'images de système de reconnaissance de plaque d'immatriculation Download PDF

Info

Publication number
WO2020076264A2
WO2020076264A2 PCT/TR2019/050500 TR2019050500W WO2020076264A2 WO 2020076264 A2 WO2020076264 A2 WO 2020076264A2 TR 2019050500 W TR2019050500 W TR 2019050500W WO 2020076264 A2 WO2020076264 A2 WO 2020076264A2
Authority
WO
WIPO (PCT)
Prior art keywords
tespit
bir
yolcu
sürücü
modeli
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/TR2019/050500
Other languages
English (en)
Other versions
WO2020076264A9 (fr
WO2020076264A3 (fr
Inventor
Emre BASESKI
Burak BALCI
Yusuf Oğuzhan ARTAN
Alperen ELIHOS
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.)
Havelsan Hava Elektronik Sanayi ve Ticaret AS
Original Assignee
Havelsan Hava Elektronik Sanayi ve Ticaret 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 Havelsan Hava Elektronik Sanayi ve Ticaret AS filed Critical Havelsan Hava Elektronik Sanayi ve Ticaret AS
Publication of WO2020076264A2 publication Critical patent/WO2020076264A2/fr
Publication of WO2020076264A9 publication Critical patent/WO2020076264A9/fr
Publication of WO2020076264A3 publication Critical patent/WO2020076264A3/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
    • G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
    • G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules

Definitions

  • Bu yolla sunucuda [107] toplanan görüntüler bir haf ⁇ za eleman ⁇ nda toplanarak görüntü kümesi [201] oladosturulur.
  • Burada rotedan [102] gelen tüm görüntüler mevcuttur.
  • Bu görüntü kümesi e ⁇ itim görüntüleri [202] ve test görüntüleri [211] olarak iki gruba ayr ⁇ l ⁇ r.
  • Buna göre e ⁇ itim görüntüleri [202] kullan ⁇ larak ön cam bölgesi tespit modeli oladosturulmas ⁇ [203] i warlemi tetiklenir.
  • Burada olavaturulan ön cam tespit modeli [301] sunucu üzerinde saklan ⁇ r.
  • Bu modellerden biri olan SSD (tek seferde 159%) kutu taramas ⁇ /single shot multibox detection), 159klu evri->->When a ⁇ lar ⁇ (multiple evolutionary neural networks) üzerinde tespit edilecek objeleri görsel ve yap ⁇ sal özelikleri ile temsil etmeyi ö ⁇ renmektedir.
  • Üçüncü ad ⁇ mda ise bu ayr ⁇ lan görüntüler yolcu kemer tespit modeli [303] ve sürücü kemer tespit modeli [304] üzerinde i warlenerek yolcu kemer ihlal tespit sonucu [217] ve sürücü kemer ihlal tespit sonucu [218] verisine dönüwtmaschine.
  • Buna göre bir karayolu [100] üzerinde seyreden bir araç [101] k ⁇ z ⁇ lötesi ⁇ w ⁇ k kayna ⁇ [103] taraf ⁇ ndan ayd ⁇ nlat ⁇ lan alanda bulundu ⁇ unda sch tanaut ⁇ y ⁇ c ⁇ platform [104] üzerine yerlinstitutrilmikos en az bir ische [102] görücan alan ⁇ na girer ve en az bir görüntüsü al ⁇ n ⁇ r. Bu görüntü -tercihen internet [105] üzerinden- sunucuya [107] aktar ⁇ l ⁇ r
  • ⁇ kincisi ise, bu seçim kesilir ve bir sonraki model olavaturma i warleminde kullan ⁇ lmak üzere sunucuda saklan ⁇ r.
  • ⁇ kincisi ise, bu seçim kesilir ve bir sonraki model olavaturma i warleminde kullan ⁇ lmak üzere sunucuda saklan ⁇ r.
  • Bu ixxaretleme tercihen dikdörtgen bir seçim ile yap ⁇ l ⁇ r; bu hem kullan ⁇ c ⁇ n ⁇ n iwararetlemesini hem de sonraki ad ⁇ mlarda yap ⁇ lan kesme ikoslemini kolaylagrot ⁇ rmak için dikdörtgen olarak belirtilmiwar olup, teknikte uzman bir ki might herhangi bir geometrik wekil seçebilir.
  • Burada ön cam bölgesi tespit modeli [301] oladosturulur ve sunucu [107] üzerinde daha sonra kullan ⁇ lmak üzere saklan ⁇ r. E ⁇ itim görüntülerinin [202] her birinin i warlemden geçirilmesinden sonra ön cam bölgesi tespit modeli [301] güncellenir.
  • Bu ixxaretleme tercihen dikdörtgen bir seçim ile yap ⁇ l ⁇ r; bu hem kullan ⁇ c ⁇ n ⁇ n iwararetlemesini hem de sonraki ad ⁇ mlarda yap ⁇ lan kesme ikoslemini kolaylagrot ⁇ rmak için dikdörtgen olarak belirtilmiwar olup, teknikte uzman bir ki might herhangi bir geometrik wekil seçebilir. Burada sürücü ve yolcu tespit modeli [302] olavaturulur.
  • Her ne kadar sürücü tespit edilmesi [205] ve yolcu tespit edilmesi [206] i souplemleri geometrik sebeplerle farkl ⁇ i souplemler olarak yap ⁇ lsa da her bir e ⁇ itim görüntüsü üzerinde art arda yap ⁇ lmalar ⁇ sebebiyle anlat ⁇ m ⁇ kolaylagrot ⁇ rmak amac ⁇ yla tek bir i souplem ad ⁇ m ⁇ olarak bahsedilmsuitr.
  • Bu ixxaretleme tercihen dikdörtgen bir seçim ile yap ⁇ l ⁇ r; bu kullan ⁇ c ⁇ n ⁇ n iwararetlemesini kolaylaformt ⁇ rmak için dikdörtgen olarak belirtilmiwar olup, teknikte uzman bir ki might herhangi bir geometrik wekil seçebilir.
  • Burada yolcu kemer tespit modeli [303] olavaturulur ve sunucu [107] üzerinde daha sonra kullan ⁇ lmak üzere saklan ⁇ r.
  • E ⁇ itim görüntülerinin [202] her birine denk gelen yolcu görüntüsünün i souplemden geçirilmesinden sonra yolcu kemer tespit modeli [303] güncellenir.
  • Bu ixxaretleme tercihen dikdörtgen bir seçim ile yap ⁇ l ⁇ r; bu kullan ⁇ c ⁇ n ⁇ n iwararetlemesini kolaylaformt ⁇ rmak için dikdörtgen olarak belirtilmiwar olup, teknikte uzman bir ki might herhangi bir geometrik wekil seçebilir.
  • Görüntüleri Görüntü kümesi [201] içinden ayr ⁇ larak e ⁇ itim görüntüleri [202] olarak belirlenmeyen görüntülerin kümesidir.
  • Cam Bölgesi Tespiti Sunucu [107] üzerinde saklanan test görüntülerinin [211] ön cam bölgesi tespit modeli [301] ile i warlenmesi sonucunda otomatik olarak gerçekle complexrilen ön cam bölgesi tespit edilmesi i vinlemidir.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Agricultural Chemicals And Associated Chemicals (AREA)
  • Cosmetics (AREA)
  • Image Analysis (AREA)
  • Traffic Control Systems (AREA)

Abstract

L'invention concerne un procédé pour assurer l'apprentissage par la formation de multiples modèles d'apprentissage profond au moyen d'une image de traitement reçue d'une caméra de système de reconnaissance de plaques d'immatriculation (LPRS), qui n'ont pas la taille, la dimension et la résolution standard avec diverses étapes de traitement, et informer des unités de sécurité sur des violations en termes de port de ceinture de sécurité au moyen de ladite technique d'apprentissage. En conséquence, les données de caméra sont reçues d'un serveur dans le poste de police via une connexion Internet et un modèle est formé dans ces images de caméra pour détecter un pare-brise; le pare-brise est détecté à l'aide de ce modèle. On a formé un modèle pour détecter un passager et un conducteur dans la zone de pare-brise; la détection du conducteur et du passager est effectuée dans la zone de pare-brise sur ledit modèle. On a formé un modèle pour détecter une ceinture de sécurité sur le conducteur et la zone de passager, et le résultat de détection de ceinture de sécurité est formé sur le modèle en fonction de valeurs de score de la détection de ceinture de sécurité sur les images d'un passager d'exemple.
PCT/TR2019/050500 2018-06-25 2019-06-25 Procédé de détection de violation de ceinture de sécurité à l'aide d'images de système de reconnaissance de plaque d'immatriculation Ceased WO2020076264A2 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
TR2018/08941A TR201808941A2 (tr) 2018-06-25 2018-06-25 Plaka tanima si̇stemi̇ kamera görüntüleri̇ kullanilarak emni̇yet kemer i̇hlal tespi̇ti̇ i̇çi̇n bi̇r yöntem
TR2018/08941 2018-06-25

Publications (3)

Publication Number Publication Date
WO2020076264A2 true WO2020076264A2 (fr) 2020-04-16
WO2020076264A9 WO2020076264A9 (fr) 2020-05-28
WO2020076264A3 WO2020076264A3 (fr) 2020-06-25

Family

ID=67001021

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/TR2019/050500 Ceased WO2020076264A2 (fr) 2018-06-25 2019-06-25 Procédé de détection de violation de ceinture de sécurité à l'aide d'images de système de reconnaissance de plaque d'immatriculation

Country Status (2)

Country Link
TR (1) TR201808941A2 (fr)
WO (1) WO2020076264A2 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2025242366A1 (fr) 2024-05-22 2025-11-27 Fareco Systèmes et méthodes de détection du port de la ceinture de sécurité par un passager d'un véhicule

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020142050A2 (fr) * 2018-12-31 2020-07-09 Havelsan Hava Elektroni̇k Sanayi̇ Ve Ti̇caret Anoni̇m Şi̇rketi̇ Procédé pour détecter l'infraction d'une utilisation de téléphone mobile par un conducteur au moyen d'images de caméra du système de reconnaissance de plaque d'immatriculation

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8824742B2 (en) * 2012-06-19 2014-09-02 Xerox Corporation Occupancy detection for managed lane enforcement based on localization and classification of windshield images
US9721173B2 (en) * 2014-04-04 2017-08-01 Conduent Business Services, Llc Machine learning approach for detecting mobile phone usage by a driver
US9552524B2 (en) * 2014-09-15 2017-01-24 Xerox Corporation System and method for detecting seat belt violations from front view vehicle images

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2025242366A1 (fr) 2024-05-22 2025-11-27 Fareco Systèmes et méthodes de détection du port de la ceinture de sécurité par un passager d'un véhicule
FR3162545A1 (fr) 2024-05-22 2025-11-28 Fareco Systèmes et méthodes de détection du port de la ceinture de sécurité par un passager d’un véhicule

Also Published As

Publication number Publication date
WO2020076264A9 (fr) 2020-05-28
WO2020076264A3 (fr) 2020-06-25
TR201808941A2 (tr) 2018-09-21

Similar Documents

Publication Publication Date Title
CN107851195B (zh) 利用神经网络进行目标检测
US11257220B2 (en) Asset tracking systems
US20210157912A1 (en) Defending machine learning systems from adversarial attacks
CN110321916A (zh) 学习数据制作方法、学习方法、危险预测方法及装置、程序存储介质
EP3523749B1 (fr) Détection et classification d'objet avec des ventilateurs de fourier
US9928424B2 (en) Side window detection through use of spatial probability maps
Yazi et al. Classification of methamorphic malware with deep learning (LSTM)
CN106650644B (zh) 驾驶员危险行为识别方法及系统
CN107978110A (zh) 基于图像匹配的电子围栏智能在位识别系统及识别方法
KR101736173B1 (ko) 관심 객체 고속 검출 장치 및 그 방법
WO2022027893A1 (fr) Procédé et appareil de détection de port de ceinture de sécurité, dispositif électronique, support de stockage et programme
WO2020076264A9 (fr) Procédé de détection de violation de ceinture de sécurité à l'aide d'images de système de reconnaissance de plaque d'immatriculation
CN113900101A (zh) 障碍物检测方法、装置及电子设备
CN116092333B (zh) 航班保障节点的确定方法、装置、电子设备及存储介质
US20120155711A1 (en) Apparatus and method for analyzing video
CN113408364A (zh) 一种临时车牌识别方法、系统、装置及存储介质
CN110705416B (zh) 一种基于驾驶员面部图像建模的安全驾驶预警方法及系统
Cyganek et al. Vehicle logo recognition with an ensemble of classifiers
CN109960993A (zh) 车辆安全检测系统及方法、电子设备、存储介质
EP3572970A1 (fr) Système de sécurité et procédé pour entraînement autonome et/ou d'aide à la conduite dans un véhicule à moteur
CN112232317B (zh) 用于目标朝向识别的目标检测方法及装置、设备及介质
CN107334481B (zh) 一种驾驶分心检测方法及系统
WO2020046252A2 (fr) Procédé de détection de violation de capacité dans un siège passager avant faisant appel à des images d'un système de reconnaissance de plaque d'immatriculation
Wang et al. Vehicle identification via sparse representation
US20180201261A1 (en) Method for checking the plausibility of a control decision for safety means

Legal Events

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

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19870150

Country of ref document: EP

Kind code of ref document: A2