非物质文化遗产视频语义标注方法研究

Research on the Method of Semantic Annotation for Intangible Cultural Heritage Videos

【目的/意义】针对非物质文化遗产(简称"非遗")视频资源难以入库与管理的难题,提出了一种非遗视频资源的管理和语义组织的语义标注方法。【方法/过程】首先对非遗视频采用基于SURF算法和窗口阈值的镜头分割方法,将非遗视频解构成不同颗粒度的逻辑单元;其次构建了非遗视频语义标注的本体模型;最后通过视频语义标注过程,实现领域知识概念与视频资源间的语义关联,赋予视频单元丰富的语义信息。【结果/结论】非遗视频综合语义标注方法可对非遗视频资源实现有效管理,为用户提供结构化、语义化的视频浏览及知识可视化等知识服务,对非遗数字化保护与传播具有重要的意义和价值。

【Purpose/significance】In this paper, a semantic annotation method for management and semantic organizationis proposed to solve the difficult problems of storage and management of Intangible Cultural Heritage video resources.【Method/process】Firstly, the video is partitioned into a logical unit with different granularity using the SURF algorithmand threshold segmentation method. Secondly, the paper constructed the ontology model of semantic annotation forIntangible Cultural Heritage videos. Finally, the semantic association between the domain knowledge concept and the videoresource is realized through the video semantic annotation process, which gives the video unit abundant related semanticinformation.【Result/conclusion】The comprehensive video semantic annotation method can manage the video resourceseffectively and provide users with knowledge visualization and knowledge services such as structured and semantic videobrowsing. Moreover, it is of great significance and value for digital preservation and dissemination of Intangible CulturalHeritage.