苏科华

Supervisor of Doctorate Candidates  
Supervisor of Master's Candidates

E-Mail:

Date of Employment:2008-11-02

School/Department:计算机学院

Education Level:研究生毕业

Business Address:D203

Gender:Male

Contact Information:13517299596

Status:Employed

Discipline:Computer Applications Technology
Communications and Information Systems
Other specialties in Software Engineering
Cyberspace Security


Paper Publications

Identify intersections, curves and bifurcation junctions from GPS traces

Hits:

DOI number:10.4028/www.scientific.net/AMM.253-255.1666

Affiliation of Author(s):TRANS TECH PUBLICATIONS LTD

Journal:SUSTAINABLE DEVELOPMENT OF URBAN INFRASTRUCTURE, PTS 1-3

Place of Publication:KREUZSTRASSE 10, 8635 DURNTEN-ZURICH, SWITZERLAND

Key Words:GPS;road map;cross detection;intersections;curves;bifurcation junctions

Abstract:With the increasing popularity of route guidance systems, it becomes more important for such highly advanced applications as the cross detection to induce road maps with a higher accuracy. In this paper, we present a novel approach to induce higher-precision maps with the help of the GPS data, mainly identify intersections, curves, and bifurcation junctions. Our approach consists of several successive processing steps: Firstly, try to find out areas where the distribution of GPS feature points is very intensive. Then, analyze the characteristic of the intersections and attach some restrictions to filter out the intersections, curves and bifurcation junctions which meet with our conditions. We also propose this method to improve the efficient of our algorithm. Among the new contributions are a creative method that connects both the TargetRegion_Detect algorithm and the AbnormalPointFilter algorithm which helps us identify the three forms of roads mentioned above and induce road maps with a higher accuracy.

Co-author:Zhang, Yu,Cheng, Hong

First Author:Su, Kehua

Indexed by:Article

Correspondence Author:Wang, Haofeng

Document Type:J

Volume:253-255

Page Number:1666-1674

ISSN No.:1660-9336

Translation or Not:no

Date of Publication:2013-01-01

Profile

苏科华,男,武汉大学计算机学院教授、博导;武汉大学科技成果转化中心(技术转移中心)副主任。研究主要集中在最优传输(Optimal Transport)领域,它是研究概率测度间最优变换的一类优化问题。在计算机图形学、机器视觉、人工智能、医学图像处理等领域有着广泛的应用。本人主要研究最优传输的几何计算理论和高效算法,并将其应用于网格保测参数化、三维场景优化、智能烧伤评估和卫星互联网任务优化中。主持包括国家自然科学基金、中央军科委、航天5院、华为公司等20多个项目支持,发表论文50余篇,获批发明专利10余项。为CCF计算机辅助设计与图形学(CAD/CG)和虚拟现实与可视化(TCVRV)专委会执行委员。