AUTONOMOUS DRIVING ASSISTANCE GLASSES THAT ASSIST IN AUTONOMOUS DRIVING BY RECOGNIZING HUMANS' STATUS AND DRIVING ENVIRONMENT THROUGH IMAGE ANALYSIS BASED ON DEEP NEURAL NETWORK

The present invention relates to a method for providing driving safety information through auxiliary glasses worn by a driver. The method of the present invention, which is performed by a driving safety information analyzing device, comprises: a step (a) of generating a feature map by inputting a vi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: JE HONG MO, LEE HYUNG SOO, JEONG KYUNG JOONG, JANG TAE WOONG, SHIN DONG SOO, LEE MYEONG CHUN, KIM KYE HYEON, KIM YONG JOONG, YEO DONG HUN, RYU WOO JU, NAM WOON HYUN, SUNG MYUNG CHUL, KIM HAK KYOUNG, CHO HO JIN, BOO SUK HOON
Format: Patent
Sprache:eng ; kor
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The present invention relates to a method for providing driving safety information through auxiliary glasses worn by a driver. The method of the present invention, which is performed by a driving safety information analyzing device, comprises: a step (a) of generating a feature map by inputting a vision-dependent driving image to a convolutional network when the vision-dependent driving image corresponding a vision of a driver is acquired from a camera mounted on auxiliary glasses, and acceleration information and gyroscope information from a sensor are acquired, inputting the feature map to a detection network, a segmentation network, and an identification network, respectively, so that the detection network detects an object, the segmentation network detects a lane, and the identification network detects a driving environment, and generating status information on the driver by inputting the acceleration information and the gyroscope information to a circulation network; and a step (b) of notifying the driver of predicted collision probability information, lane departure information, and the driving environment, and providing a warning. Accordingly, at least one driver status and at least one driving environment can be detected at low costs. 운전자가 착용한 보조 안경을 통해 주행 안전 정보를 제공하는 방법으로서, 주행 안전 정보 분석 장치가, (a) 상기 보조 안경에 탑재된 카메라로부터, 상기 운전자의 시각에 대응되는 시각 종속형 주행 이미지, 센서로부터의 가속도 정보 및 자이로스코프(gyroscope) 정보가 획득되면, 상기 시각 종속형 주행 이미지를 컨벌루션 네트워크로 입력하여 특징 맵을 생성하도록 하고, 상기 특징 맵을 검출 네트워크, 세그먼테이션 네트워크, 및 식별 네트워크로 각각 입력함으로써, 상기 검출 네트워크로 하여금 객체를 검출하도록 하며, 상기 세그먼테이션 네트워크로 하여금 차선을 검출하도록 하고, 상기 식별 네트워크로 하여금 주행 환경을 검출하도록 하며, 상기 가속도 정보 및 상기 자이로스코프 정보를 순환 네트워크로 입력하여 상기 운전자의 상태 정보를 생성하도록 하는 단계; 및 (b) 예측 충돌 가능성 정보, 차선 이탈 정보, 상기 주행 환경을 상기 운전자에게 알려주고, 경고를 제공하는 단계;를 포함하는 방법이 제공된다.