A Terrain Elevation Map Generation Method Based on Self-Attention Mechanism and Multifeature Sketch

Author:

Cai Xingquan1ORCID,Xi Mengyao1ORCID,Yu Nu1ORCID,Yang Zhe1ORCID,Sun Haiyan1ORCID

Affiliation:

1. School of Information Science and Technology, North China University of Technology, Beijing 100144, China

Abstract

To address the issues of low efficiency in manual terrain feature map annotating and poor realism in terrain elevation map generation, this paper proposes a terrain elevation map generation method based on self-attention mechanism and multifeature sketch. Firstly, the proposed method extracts features from a terrain elevation map using an adaptive feature enhancement method. Afterwards, our method adds a self-attention mechanism to the generator and discriminator of conditional generative adversarial network to capture the global spatial features and generates a realistic terrain elevation map. Finally, a level of detail method is used to visualize the three-dimensional terrain, and an interactive terrain editing tool for roaming interaction is implemented. Experimental data show that the proposed method performs well in subjective visual performance and objective criteria and has obvious advantages over other current typical methods.

Funder

National Natural Science Foundation of China

Publisher

Hindawi Limited

Subject

General Mathematics,General Medicine,General Neuroscience,General Computer Science

Reference24 articles.

1. Interactive generation and representation of three-dimensional terrain;H. C. Song;Computer Engineering and Applications,2004

2. Interactive digital terrain synthesis algorithm;H. F. Yin;Journal of Computer-Aided Design and Graphics,2012

3. Adaptive multi-feature fusion for fast realistic terrain mapping;Y. Y. Zeng;” Journal of Image and Graphics,2013

4. An efficient parallel method for topography generation by Perlin noise;Y. Liu;Bulletin of Science and Technology,2016

5. Interactive example-based terrain authoring with conditional generative adversarial networks;É. Guérin;ACM Transactions on Graphics,2017

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3