Matches in SemOpenAlex for { <https://semopenalex.org/work/W3197078790> ?p ?o ?g. }
Showing items 1 to 68 of
68
with 100 items per page.
- W3197078790 endingPage "1329" @default.
- W3197078790 startingPage "1323" @default.
- W3197078790 abstract "도로에는 낙하물, 포트홀, 로드킬 등 다양한 위험객체가 빈번하게 발생하며, 이로 인한 사고를 예방하기 위한 최선의 방법은 신속하게 검지하는 것이다. 이를 위해 전 세계적으로 모바일 앱 기반의 도로위험객체 신고 서비스가 공공과 민간 영역에서 활발히 개발 · 운영되고 있는 실정이다. 하지만 도로환경에서 앱을 사용하여 위험객체를 신고하는 행위는 위험하기 때문에 그 사용시간을 최소화하는 기술의 개발이 필요하다. 이에 본 연구에서는 딥러닝 기술을 활용하여 도로 이미지로부터 도로위험객체를 자동으로 인식하고 분류하는 방법을 제안 및 구현하고 성능을 검증함으로써 그 가능성을 점검하였다. 신고 이미지는 노면상태불량, 배수시설불량, 도로시설물불량, 로드킬 및 장애물의 네 개의 분류체계로 분류하였고, 딥러닝을 통해 학습된 모형(YOLO v3)은 평균 95%의 높은 검출율을 나타냈다. Various dangerous road objects such as potholes, falling objects, roadkills occur frequently on road, and the best way to prevent accidents from these dangers is to find them as soon as possible. A lot of mobile applications to report the road problems have been developed and in service worldwide both in public and private sectors. It is not safe, however, to report the problem using those apps in road environment, and it is necessary to develop a technology to minimize the time to make a report using the apps. For this purpose, a method to recognize and classify the road dangerous objects automatically from road images using deep learning algorithm was proposed and implemented, and the performance of the proposed model(YOLO v3) was tested, which shows 95% success rate on average to detect the four categories of road dangers including pavement, drainage, road facility, and roadkill, and demonstrates the possibility of the proposed method." @default.
- W3197078790 created "2021-09-13" @default.
- W3197078790 creator A5032929028 @default.
- W3197078790 creator A5053264584 @default.
- W3197078790 creator A5072729723 @default.
- W3197078790 date "2021-08-31" @default.
- W3197078790 modified "2023-10-18" @default.
- W3197078790 title "A study to recognize and classify road dangerous objects automatically using deep learning" @default.
- W3197078790 doi "https://doi.org/10.9728/dcs.2021.22.8.1323" @default.
- W3197078790 hasPublicationYear "2021" @default.
- W3197078790 type Work @default.
- W3197078790 sameAs 3197078790 @default.
- W3197078790 citedByCount "0" @default.
- W3197078790 crossrefType "journal-article" @default.
- W3197078790 hasAuthorship W3197078790A5032929028 @default.
- W3197078790 hasAuthorship W3197078790A5053264584 @default.
- W3197078790 hasAuthorship W3197078790A5072729723 @default.
- W3197078790 hasConcept C108583219 @default.
- W3197078790 hasConcept C127313418 @default.
- W3197078790 hasConcept C127413603 @default.
- W3197078790 hasConcept C144133560 @default.
- W3197078790 hasConcept C154945302 @default.
- W3197078790 hasConcept C162853370 @default.
- W3197078790 hasConcept C22212356 @default.
- W3197078790 hasConcept C2776023743 @default.
- W3197078790 hasConcept C2779079380 @default.
- W3197078790 hasConcept C2780378061 @default.
- W3197078790 hasConcept C38652104 @default.
- W3197078790 hasConcept C41008148 @default.
- W3197078790 hasConcept C5900021 @default.
- W3197078790 hasConcept C71924100 @default.
- W3197078790 hasConcept C99454951 @default.
- W3197078790 hasConceptScore W3197078790C108583219 @default.
- W3197078790 hasConceptScore W3197078790C127313418 @default.
- W3197078790 hasConceptScore W3197078790C127413603 @default.
- W3197078790 hasConceptScore W3197078790C144133560 @default.
- W3197078790 hasConceptScore W3197078790C154945302 @default.
- W3197078790 hasConceptScore W3197078790C162853370 @default.
- W3197078790 hasConceptScore W3197078790C22212356 @default.
- W3197078790 hasConceptScore W3197078790C2776023743 @default.
- W3197078790 hasConceptScore W3197078790C2779079380 @default.
- W3197078790 hasConceptScore W3197078790C2780378061 @default.
- W3197078790 hasConceptScore W3197078790C38652104 @default.
- W3197078790 hasConceptScore W3197078790C41008148 @default.
- W3197078790 hasConceptScore W3197078790C5900021 @default.
- W3197078790 hasConceptScore W3197078790C71924100 @default.
- W3197078790 hasConceptScore W3197078790C99454951 @default.
- W3197078790 hasIssue "8" @default.
- W3197078790 hasLocation W31970787901 @default.
- W3197078790 hasOpenAccess W3197078790 @default.
- W3197078790 hasPrimaryLocation W31970787901 @default.
- W3197078790 hasRelatedWork W2126887587 @default.
- W3197078790 hasRelatedWork W2731899572 @default.
- W3197078790 hasRelatedWork W2939353110 @default.
- W3197078790 hasRelatedWork W3009238340 @default.
- W3197078790 hasRelatedWork W3175064897 @default.
- W3197078790 hasRelatedWork W3215138031 @default.
- W3197078790 hasRelatedWork W4312962853 @default.
- W3197078790 hasRelatedWork W4321369474 @default.
- W3197078790 hasRelatedWork W4327774331 @default.
- W3197078790 hasRelatedWork W4360585206 @default.
- W3197078790 hasVolume "22" @default.
- W3197078790 isParatext "false" @default.
- W3197078790 isRetracted "false" @default.
- W3197078790 magId "3197078790" @default.
- W3197078790 workType "article" @default.