ODOMATIC: Automatic Species Identification, Functional Morphology, and Feature Extraction to alleviate the taxonomic impediment and broaden citizen science tools.

生物体的命名和分类是生物学建立的一项基本原则。事实上，我们生活的几乎方方面面--从我们吃的食物到我们吃的药，我们住的房子，到我们欣赏的自然风景--几乎都与地球上存在的生物多样性有关。所有物种都必须被命名和编目，以便人类了解我们周围的世界。近几十年来，负责识别和描述物种的分类学科学领域的专家队伍有所减少；这是不幸的，因为随着人类活动导致生物多样性前所未有的下降，对分类学的需求从未像现在这样强烈。该项目将通过引进一套开放源码的、基于网络的工具，从图像中识别物种，并为比较研究测量成像标本，从而帮助减轻当前和未来的分类学家在识别和编目生物多样性方面的负担。这些软件工具将通过公共资助的CyVerse网络平台免费提供。此外，一个名为ODOMATIC的识别蜻蜓和蜻蜓的系统将在现有的OdonataCentral网站上实现，使研究人员和爱好者能够从它们翅膀的图像中准确地识别这些昆虫。将在新泽西州和阿拉巴马州的不同城市社区提供一系列免费研讨会，以增加OdonataCentral？S用户基础的多样性，鼓励参与STEM领域，并总体上增加对自然世界的欣赏和理解。一系列Google Hangout活动将鼓励拉丁美洲、非洲和亚洲的世界蜻蜓协会成员为ODOMATIC提供国际用户。此外，纽瓦克罗格斯大学的本科生将被招募来帮助设计和培训ODOMATIC，给他们在研究、编程和分类学方面的宝贵经验。这项工作将有助于减少分类学障碍--迫切需要更多的分类学产品，如物种描述和标本鉴定，来自更少的分类学家工作者--通过解决耗时但必要的任务--物种鉴定。该项目的目标包括发布一种新开发的系统，用于从蜻蜓(蜻蜓和蜻蜓)的翅膀图像中自动识别它们，该系统使用计算机视觉和机器学习来描述物种和对物种进行分类。将部署一个界面，使使用该系统的物种鉴定可以通过OdonataCentral网站(odonatacentral.org)进行访问。此外，还将开发独立的工具，用于在生物图像中的标本上自动描述和放置几何形态地标，以用于基于形态的比较研究。这些目标的成功完成将使牙医界受益，因为它将提供蜻蜓和蜻蜓的鉴定。更广泛地说，这里创造的工具将允许生物学家研究其他生物群体，从他们的标本图像中快速提取形态数据。

The naming and classifying of living organisms is a fundamental principal upon which biology is founded. Indeed, almost every facet of our lives - from the food we eat to the medicine we take, the houses we live in to the natural scenery we admire - has a connection to the biodiversity present on Earth. All species must be named and cataloged in order for humans to understand the world around us. The scientific field of taxonomy, responsible for identifying and describing species, has incurred a reduction in its specialist workforce in recent decades; this is unfortunate because the need for taxonomy has never been greater as human activities result in unprecedented decline in the diversity of life. This project will help to reduce the burden placed on current and future taxonomists to identify and catalog biodiversity by introducing a set of open-source, web-based tools for identifying species from images and for measuring imaged specimens for comparative studies. These software tools will be made freely available through the publicly-funded CyVerse web platform. In addition, a system for identifying dragonflies and damselflies called ODOMATIC will be implemented on an existing web presence, OdonataCentral, allowing researchers and enthusiasts to accurately identify these insects from images of their wings. A series of free workshops will be offered in diverse urban communities in New Jersey and Alabama to increase the diversity of OdonataCentral?s user base, encourage participation in the STEM fields, and generally increase appreciation and understanding of the natural world. A series of Google Hangout events will encourage international user-ship for ODOMATIC from World Dragonfly Association members in Latin America, Africa, and Asia. In addition, undergraduate students from Rutgers University-Newark will be recruited to help in designing and training ODOMATIC, giving them valuable experience in research, programming and taxonomy. This work will help to reduce the taxonomic impediment - an urgent need for more taxonomic products, like species descriptions and specimen identifications, from fewer taxonomist workers - by addressing the time-consuming, but necessary, task that is species identification. Objectives of this project include release of a newly-developed system for automatically identifying Odonata (dragonflies and damselflies) from images of their wings, which uses computer vision and machine learning to characterize and classify species. An interface will be deployed making species identification using this system accessible through the OdonataCentral website (odonatacentral.org). In addition, stand-alone tools will be developed for automatically describing and placing geometric morphometric landmarks on specimens in biological imagery for use in morphology-based comparative studies. The successful completion of these objectives will benefit the odonatological community by providing dragonfly and damselfly identification. More broadly, the tools created here will allow biologists studying other groups of organisms to rapidly extract morphological data from images of their specimens.