基于内容的遥感图像变化信息检索概念模型设计

Design of Content-based Remote Sensing Image Change Information Retrieval and Relevance Feedback Model

在海量遥感数据背景下,传统的基于关键字/元数据数据服务模式,无法满足不同应用领域用户对多样化遥感变化信息数据的获取需求。将基于内容的图像检索技术应用到遥感图像变化信息数据获取中,提出了一种全新的基于内容的遥感图像变化信息检索概念模型。通过深入分析当前基于内容的图像检索的先进理论方法,构建基于内容的遥感图像变化信息检索模型框架,并对变化信息数据管理模型构建、多维特征提取和智能反馈模型创建等关键问题进行研究和算法实现,以中低分辨率遥感图像变化信息数据获取为例来进行模型验证与分析,建立原型系统。该方法作为一种新的遥感图像变化信息获取与服务方式,能有效利用遥感图像中底层特征,更准确地刻画了不同用户的遥感图像变化信息检索需求。同时,对影像的预处理要求较低,不受变化检测产品生产种类限制,具有较好普适性和自动化性,提高了遥感信息服务水平和效率。

With the rapid development of satellite remote sensing technology,processing the variety of remotely sensed data has increasingly been complex and difficult. It is also hard to efficiently and intelligently retrieve change information what users need from a massive database of images. In the context of mass remote sensing data,the existing knowledge based on a priori knowledge + the keyword/metadata remote sensing data service model can not meet above-mentioned challenge. Firstly,it is not guaranteed to obtain the totally change information data in the database,as we can not get the all prior knowledge. Second,the keyword/metadata can not accurately describe the different application areas of the user’s actual retrieval needs. To deal with this,the Content-Based Image Retrieval(CBIR) is successfully applied on the change detection in this paper. And,Content-Based Remote Sensing Image Change Information Retrieval and Relevance Feedback model is introduced. Firstly,we learn the CBIR theory fully and exclusively. Then,the model structure and framework is built. And,some critical issues,such as data management,multi-features selection and relevance feedback,are considered. Thirdly,an experimental prototype system is set up to demonstrate the validity and practicability of this model. As a new remote sensing image change detection information acquisition mode,the new model can reduce the demands of image pre-processing,overcome synonyms spectrum,seasonal changes and other factors in the change detection,and meet different kinds of needs. Meanwhile,the new model has important implications for improving remote sensing image management skill and autonomic capabilities of information retrieval filed.