• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권8호
• /
• pp.1996-2015
• /
• 2023

High-definition (HD) maps can provide precise road information that enables an autonomous driving system to effectively navigate a vehicle. Recent research has focused on leveraging semantic segmentation to achieve automatic annotation of HD maps. However, the existing methods suffer from low recognition accuracy in automatic driving scenarios, leading to inefficient annotation processes. In this paper, we propose a novel semantic segmentation method for automatic HD map annotation. Our approach introduces a new encoder, known as the convolutional transformer hybrid encoder, to enhance the model's feature extraction capabilities. Additionally, we propose a multi-level fusion module that enables the model to aggregate different levels of detail and semantic information. Furthermore, we present a novel decoupled boundary joint decoder to improve the model's ability to handle the boundary between categories. To evaluate our method, we conducted experiments using the Bird's Eye View point cloud images dataset and Cityscapes dataset. Comparative analysis against stateof-the-art methods demonstrates that our model achieves the highest performance. Specifically, our model achieves an mIoU of 56.26%, surpassing the results of SegFormer with an mIoU of 1.47%. This innovative promises to significantly enhance the efficiency of HD map automatic annotation.

• International Journal of Automotive Technology
• /
• 제7권3호
• /
• pp.361-367
• /
• 2006

This paper describes a novel stereo vision based localization of free parking site, which recognizes the target position of automatic parking system. Pixel structure classification and feature based stereo matching extract the 3D information of parking site in real time. The pixel structure represents intensity configuration around a pixel and the feature based stereo matching uses step-by-step investigation strategy to reduce computational load. This paper considers only parking site divided by marking, which is generally drawn according to relevant standards. Parking site marking is separated by plane surface constraint and is transformed into bird's eye view, on which template matching is performed to determine the location of parking site. Obstacle depth map, which is generated from the disparity of adjacent vehicles, can be used as the guideline of template matching by limiting search range and orientation. Proposed method using both the obstacle depth map and the bird's eye view of parking site marking increases operation speed and robustness to visual noise by effectively limiting search range.

• 대한임베디드공학회논문지
• /
• 제17권1호
• /
• pp.41-49
• /
• 2022

This paper proposes a novel lane detection algorithm based on inverse perspective transformation and machine learning in lightweight embedded system. The inverse perspective transformation method is presented for obtaining a bird's-eye view of the scene from a perspective image to remove perspective effects. This method requires only the internal and external parameters of the camera without a homography matrix with 8 degrees of freedom (DoF) that maps the points in one image to the corresponding points in the other image. To improve the accuracy and speed of lane detection in complex road environments, machine learning algorithm that has passed the first classifier is used. Before using machine learning, we apply a meaningful first classifier to the lane detection to improve the detection speed. The first classifier is applied in the bird's-eye view image to determine lane regions. A lane region passed the first classifier is detected more accurately through machine learning. The system has been tested through the driving video of the vehicle in embedded system. The experimental results show that the proposed method works well in various road environments and meet the real-time requirements. As a result, its lane detection speed is about 3.85 times faster than edge-based lane detection, and its detection accuracy is better than edge-based lane detection.

• 한국정보기술학회 영문논문지
• /
• 제10권2호
• /
• pp.167-175
• /
• 2020

In this paper, we propose a traffic accident detection on vehicle-mounted camera. In the proposed method, the minimum bounding box coordinates the central coordinates on the bird's eye view and motion vectors of each vehicle object, and ego-motions of the vehicle equipped with dash-cam are extracted from the dash-cam video. By using extracted 4 kinds features as the input of Bi-LSTM (bidirectional LSTM), the accident probability (score) is predicted. To investigate the effect of each input feature on the probability of an accident, we analyze the performance of the detection the case of using a single feature input and the case of using a combination of features as input, respectively. And in these two cases, different detection models are defined and used. Bi-LSTM is used as a cascade, especially when a combination of the features is used as input. The proposed method shows 76.1% precision and 75.6% recall, which is superior to our previous work.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2020년도 제62차 하계학술대회논문집 28권2호
• /
• pp.71-72
• /
• 2020

본 논문에서는 자동차 블랙박스를 사용하여 촬영된 비디오에서 자동차 사고 발생 여부를 판단하는 방법을 제안한다. 제안한 방법은 우선 객체 추적 과정에서 구한 조감도 좌표를 사용하여 각 차량 사이의 거리에 기반을 두고 교통사고 여부를 판단한다. 그런데 거리만을 사용하여 사고 여부를 판단하는 경우 자동차가 밀집된 주·정차 환경에서는 오검출의 확률이 높아질 수 있다. 이를 위해 각 차량에 대한 움직임 벡터를 계산하고 벡터 간의 정보(사잇각과 크기 등)를 사용하여 차량의 주·정차 여부를 판단한 후 사고 검출 대상에서 배제할 수 있도록 한다. 주·정차 판단 여부를 통해 사고 검출의 정확도를 향상할 수 있는 것을 실험적으로 확인하였다.

• 정보처리학회논문지A
• /
• 제10A권2호
• /
• pp.165-172
• /
• 2003

본 논문에서는 프로젝션 테이블(the projection table)과 프로젝션 월(the projection wall)을 이용하여 수평 및 수직형 가상 공간을 구축하고, 이러한 가상 환경에서 삼차원 건물을 사용자의 시점에 따라 입체적으로 가시화하는 시스템을 구현하였다. 수평형 디스플레이 장비인 프로젝션 테이블은 위에서 내려다 보는 형태의 뷰잉 프러스텀(viewing frustum)을 갖기 때문에 책상이나 테이블에서의 작업을 재현하거나 bird-eye view가 필요한 응용분야에서 효과적으로 사용되어진다. 반면에 수직형 디스플레이 장비인 대형 프로젝션 월은 정면에서 바라보는 형태의 뷰잉 프러스텀을 갖기 때문에 가상 공간을 네비게이션 할 때 보다 효과적으로 사용되어진다. 본 논문에서는 주요 객체는 세밀한 삼차원 모델로 생성하고 배경은 영상을 이용함으로써 사용자와 가상 객체간의 빠른 상호작용을 제공하도록 하였으며, 모델과 영상이 합성된 결과를 상이한 가상환경에서 사용자의 위치 및 시점과 연동하여 입체적으로 보여줌으로써 사용자의 현실감을 증진시키도록 하였다.

• 한국항만학회지
• /
• 제5권2호
• /
• pp.39-63
• /
• 1991

A PC-based ship maneuvering simulator was developed which was configured in a high performance IBM PC compatible i486 and i286 computer with a TMS 340 graphic signal processor and 10 MBPS Ethernet Cards. A real-time ship maneuvering simulation program was developed which includes computer generated imagery (CGI) for bird's eye view type and perspective view type. The simulator H/W was designed and manufactured and S/W for interface of various navigation equipments was made Especially, programs for output, analysis, and assessment of simulations results were developed. Communications between PC's are made by using Ethernet bus type LAN system. Simulations could be performed under various environments (current, wind, wave etc.) using data base of harbors and ships. This system can be used for various purposes such as crew's training, harbor and waterway design, and assessment of ship maneuverability in harbor.

• 한국실내디자인학회논문집
• /
• 제22권1호
• /
• pp.105-114
• /
• 2013

This study investigated the application and the change of various architectural models by analyzing expression viewpoint media, which were applied to the visual information of digitalized 3D contemporary architectural models. The purpose of this study was to specify how modern architects have changed 3D architectural models to conceptual, logical, and formational visual information in the process of design. This study discovered a framework of analyses by theoretically investigating a relationship between expression media and expression change in the process of visualizing architectural models. Using the framework of analyses, this study analyzed how the expression viewpoints of architectural model information have been changed and applied. The transformation media of the visual information of digitalized 3D architectural models can be classified into conceptual, analytical, and formational information: 1) Contemporary architects used author-centered subjective viewpoints to express architectural concepts, which were generated in the process of their design. They selected a perspective viewpoint and a bird's eye view in order to present their architectural concepts and to depict them with one architectural model by expanding the visual scope of conceptual information. 2) Contemporary architects adopted observer-centered objective bird's eye view expression media to effectively present their architectural information to building owners and viewers. They used transformal media, which integrate architectural information into 3D and change it to different scales, in order to express their architecture logically. 3) Contemporary architects delivered model information about the generation and change of forms by expressing the image of a project from an author-centered viewpoint, instead of objectively defining formational information. They explained the generation principle of architectural forms via transformal media which develop and rotate an architectural model.

• 디지털융복합연구
• /
• 제14권12호
• /
• pp.231-236
• /
• 2016

현재 인터넷으로 감시 카메라를 이용하여 교통량을 보여주는 서비스가 제공되고 있으나, 일반적으로 사용자가 지도 위의 일정 위치를 지정하면 카메라가 장착되어 있는 기준으로 영상을 보여준다. 본 논문에서는 상단이 북쪽으로 표시되는 일반적인 지도에 현재 카메라에서 촬영되고 있는 도로의 영상을 지도 기준으로 위에서 바라보는 시점(Bird's-eye view)으로 변형(Warping)하여 보여줌으로써 사용자가 직관적으로 현재의 도로 상황을 한눈에 볼 수 있는 시스템을 제안하였다. 본 논문에서는 카메라 영상에서 보여주고 있는 도로 평면을 강인하게 추정하기 위해 밝기 변화에 강인할 수 있는 움직임 벡터를 이용하여 도로 평면을 추정하는 방법을 제안하며, 지도와 같은 방향성을 가질 수 있도록 재조정하는 과정을 적용하고 지도 위에 표시할 수 있는 사용자 친화적인 인터페이스를 제시하였다. 실험결과를 통해 본 논문에서 제안하는 방법이 지도와 같이 사용자가 인지하기 편리한 영상으로 변형되는 결과를 확인하였다.

• 한국항해학회지
• /
• 제25권4호
• /
• pp.443-453
• /
• 2001

선박조종시뮬레이터는 해기사의 교육 훈련, 항만 수로 설계 시 안전성 평가, 선박설계시 조종성능의 검토등으로 널리 활용되고 있다. 본 논문은 최근 한국해양대학교에서 개발한 선박조종시뮬레이터를 소개하고 개발 과정과 활용에 대하여 논의한다. 본 시뮬레이터는 Operation Panel, Instructor's Console, Ship Dynamics Calculation, 3D Bridge View, 2D Bird's Eye View 및 Navigational Indicators의 6구성요소로 이루어져 있으며, 이를 위해 8대의 퍼스널 컴퓨터가 배치되어 있다. 모든 구성요소들은 효율적인 정보 교환을 위하여 분산처리네트워크 방식으로 연결되어 있다. 또한, 본 논문은 항만내에서의 저속 시 조종운동 수학모델과 가상현실 모델링에 대해서도 논의한다. 마지막으로, 부산항에 대한 2축2타선박의 접안 조종 시뮬레이션 예를 보여주고 있다.