文化遗产图像检索系统实现及关键技术研究

    Implementation of Cultural Heritage Image Retrieval and Research on Its Key Techniques

    • 摘要: 为了满足人们查询文化遗产图像的需求,本文设计并实现了北京文化遗产图像检索系统(CHIR).系统充分地考虑了文化遗产图像的特点,如分类特征明显、颜色分块特征明显等.对核心的分块颜色直方图算法进行了改进,并将基于关键字、类别、颜色直方图、改进分块颜色直方图及形状特征的检索方法相结合.系统不仅具有很好的检索效果,而且具有友好的人机交互界面、快速高效的检索方式,能够很方便地浏览检索结果.文中给出了整个检索系统的架构、实现方法和系统实现的功能.

       

      Abstract: In order to fulfill people's needs of searching for cultural heritage images, the paper presents the design and implementation of Beijing Cultural Heritage Image Retrieval System. It is a unique system as the authors fully consider characters of Beijing cultural heritage image. For example, every image belongs to a specific class, many of them have rectangular color, etc. Authors improved the core method-substation color histogram method, practised the best way to combine feature extraction methods and image retrieval methods that are based on both keywords and image content together. The system is designed not only with acceptable result, but also with friendly user interface, high efficiency and convenient way to view the results. This paper also provides system architecture, implementation techniques and functions of CHIR.

       

    /

    返回文章
    返回