Computer Vision Coral Ecology: Cyber-Enabled Image Classification for Rapid, Large Scale, Automated Monitoring of Climate Change Impacts on Coral Reefs

计算机视觉珊瑚生态学：网络图像分类，用于快速、大规模、自动监测气候变化对珊瑚礁的影响

• 批准号: 0941760
• 负责人: B. Gregory Mitchell
• 金额: $ 199.94万
• 依托单位: University of California-San Diego Scripps Inst of Oceanography
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-12-01 至 2013-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0941760&HistoricalAwards=false
• 关键词: Computer; Vision; Coral; Ecology; Cyber

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).Anthropogenic stresses to reefs coupled with climate change impacts are causing unprecedented declines in coral reefs globally yet automated technology to monitor the health of coral reefs at large temporal and spatial scales does not exist. This project will develop and deploy automated technology to monitor the health of coral reefs at large temporal and spatial scales. It will create a cyber-enabled Computer Vision Coral Ecology Portal that will allow upload, and expert classification remotely via an interactive web-based server paradigm where the external user need not manage the hardware or software. To achieve this goal new computer vision technology is needed. The effort has the potential to not only advance computer vision technology but also transform the way that coral reef ecology is performed. Fundamental computer vision problems in underwater imaging and illumination modeling, registration of RGB and fluorescence images from a moving camera rig will be addressed. Progressive improvement in the computer vision system combined with new data collected with these enabling technologies, will allow for transformational ecological outcomes including: 1) creation of baselines of coral reef health at diverse locations essential for understanding impacts of climate change; 2) development of rapid methods for quantitative surveying in support of ecology research or management; 3) estimates of growth, mortality, and recruitment rates and competitive abilities of the key coral reef species and the changes associated with warming and acidification; 4) improved knowledge of deep coral reefs for which very little is known; 5) rapid and spatially extensive in situ classification in support work by our collaborators on using airborne and satellite remote sensing and 6) creation of novel methods for multi-spectral image acquisition and processing to improve cyber-based ecological classification of coral reef communities.The project includes participation and education for undergraduate students, graduate students and postdoctoral researchers. The project also includes an interdisciplinary education and out-reach program to involve high school students and teachers, using an innovative web-based "Coral Quiz" that will communicate details of ecology, computer science and the biology of coral reefs.

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。人类对珊瑚礁的压力加上气候变化的影响正在导致全球珊瑚礁前所未有的下降，但在大的时间和空间尺度上监测珊瑚礁健康的自动化技术还不存在。该项目将开发和部署自动化技术，以在大的时间和空间尺度上监测珊瑚礁的健康状况。 它将创建一个网络化的计算机视觉珊瑚生态门户网站，允许通过交互式基于网络的服务器模式远程上传和专家分类，外部用户无需管理硬件或软件。为了实现这一目标，需要新的计算机视觉技术。这项工作不仅有可能推动计算机视觉技术的发展，而且还可能改变珊瑚礁生态的执行方式。 基本的计算机视觉问题，在水下成像和照明建模，注册的RGB和荧光图像从移动摄像机钻机将得到解决。计算机视觉系统的逐步改进，结合利用这些使能技术收集的新数据，将实现转型的生态成果，包括：1）在不同地点创建珊瑚礁健康基线，这对理解气候变化的影响至关重要; 2）开发快速定量调查方法，以支持生态研究或管理;（3）对关键珊瑚礁物种的生长、死亡率、补充率和竞争能力以及与变暖和酸化有关的变化的估计;（4）增进对知之甚少的深海珊瑚礁的了解; 5）快速和空间广泛的原位分类，支持我们的合作者使用航空和卫星遥感的工作，6）创造新的方法，光谱图像采集和处理，以改进基于网络的珊瑚礁群落生态分类，该项目包括本科生、研究生和博士后研究人员的参与和教育。 该项目还包括一个跨学科教育和外联方案，让高中学生和教师参与，利用一个创新的基于网络的“珊瑚测验”，交流生态学、计算机科学和珊瑚礁生物学的细节。