视觉搜索是信息检索领域的前沿课题,它有利于揭示视觉大数据资源价值的产生机理,丰富数字图书馆知识服务思想和未来发展框架。文章围绕视觉搜索发展历程、发展趋势、概念和特点等基础问题展开分析,深入探讨视觉搜索的应用及组织模式,并归纳了视觉搜索应用与组织研究的5个核心问题。视觉搜索在国内未广泛应用的主要原因:技术与应用产品不成熟、搜索性能不稳定、用户体验质量不佳、应用局限性较强等。建议从视觉大数据资源的获取与组织、视觉大数据资源的理解与表达、视觉大数据资源整合与交互、视觉对象知识库建设及标准化、视觉搜索体系的安全与可靠性5个方面来解决。 Visual search is the forefront topic in the field of information retrieval. It is helpful to reveal the generation mechanism of the value of visual big data resources and enrich the thought and future development framework of digital library knowledge service. The article analyzes the basic problems of visual search development, such as its development trend, concept and characteristics, discusses the application and organization mode of visual search in depth, and summarizes the five core issues of visual search application and organizational research. The main reason that visual search is not widely used in China is: immature technology and application products, unstable search performance, poor user experience quality and strong application limitations. It is proposed to solve this problem from five aspects: the acquisition and organization of visual big data resources, the understanding and expression of visual big data resources, the integration and interaction of visual big data resources, the construction and standardization of visual object knowledge base and the security and reliability of visual search system.