Digitisation / Cataloguing of non-textual objects: Reference system of bioacoustic data
Digitisation / Cataloguing of non-textual objects: Reference system of bioacoustic data
批准号:
203172118
负责人:
Dr. Karl-Heinz Frommolt
金额:
$0.0万
依托单位国家:
德国
项目类别:
Cataloguing and Digitisation (Scientific Library Services and Information Systems)
财政年份:
2012
资助国家:
德国
项目状态:
已结题
起止时间:
2011-12-31 至 2014-12-31
中文摘要
野生动物声音集不仅是动物学研究的重要工具,特别是在动物行为和分类学领域,而且广泛用于音乐学问题或作为展览中声音装置的来源。近年来,生物多样性研究中的现代声学方法需要有充分的文献记录和验证的参考声音。然而,目前的声音收集仍然不能作为生物多样性研究的有效工具。通常,声音档案中的录音由长的声音序列组成,其中大部分包含背景噪音。参考资料库应包含清晰分离的声音单元,包括动物的典型歌声和/或叫声。该项目的目的是在柏林自然博物馆动物声音档案的数字化录音基础上,结合其他来源和专家,创建这样一个欧洲动物物种声音参考资料库。在参考系统中,仅包括具有高声学质量的记录。物种的确定必须可靠,记录的位置必须已知。基于网络的界面将允许无限制地访问数据库。声学记录和元数据将在该项目的网站、GBIF门户网站和Europeana上免费提供。除了通过元数据进行搜索外,还将包括基于内容的搜索。将开发和实施用于物种确定的声学模式识别算法。该系统的应用旨在用于生物多样性研究和保护中的广泛任务。此外,由于其独特性,参考系统可以指出动物发声知识的空白,并鼓励科学家通过有针对性的研究填补这些空白。
英文摘要
Wildlife sound collections are not only an important tool for zoological research, especially in the fields of animal behaviour and taxonomy, but are widely used for musicological issues or as source for sound installations in exhibitions. Recently modern acoustic methods in biodiversity research require well documented and verified reference sounds. However, current sound collections could still not serve as an effective tool for biodiversity research. Typically recordings in sound archives consist of long sound sequences mostly containing background noises. A reference library should contain clearly separated sound units including typical songs and/or calls of animals.The aim of this project is to create such a reference library of voices of European animal species on the basis of already digitized recordings of the Animal Sound Archive of the Museum für Naturkunde Berlin, integrating other sources and experts. In the reference system only records will be included that have a high acoustic quality. The determination of the species must be reliable and the location of the recording must be known. A web-based interface will allow access the database without restrictions. The acoustic records and metadata will be made freely available on the website of the project, the GBIF-Portal and Europeana. In addition to searching via metadata, a content-based search will be included. Acoustic pattern recognition algorithms for species determination will be developed and implemented. The application of the system is intended for a wide range of tasks in biodiversity research and conservation. In addition due to its encyclopaedic character the reference system can indicate gaps in knowledge on animal vocalisations and encourage scientists to fill these gaps by targeted research.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
海外基金