一种基于区域的图像检索方法

【作者】 周招娣；

【导师】 章毅；

【作者基本信息】 电子科技大学 ， 计算机软件与理论， 2010， 硕士

【摘要】 随着因特网和计算机视觉技术的发展,数字图像的产生、存储、分析和传输访问的数量呈指数级增长。全球每天都在不断地产生数以兆字节计数字图像数据,然而这些数据散乱地分布在全世界各个角落,人们不能进行有效地访问和利用。这就要求有一种能够快速而且准确地查找、访问图像的技术,也就是图像检索技术。基于内容图像检索技术,从一定程度上解决了传统的基于文本图像检索技术的局限性,已经成为当前图像检索技术的热点研究领域。基于区域图像检索技术是基于内容图像检索技术的一个重要研究方向,它从对象层次上理解和表示图像,在一定程度上缩小了图像底层特征和高层语义间的鸿沟,更符合人类视觉对图像的认知。模糊数学是研究和处理不明确现象的一种数学理论和方法。人们经常接受模糊语言与模糊信息,并能做出正确的识别和判断,因此在图像检索中结合模糊数学理论成为基于内容图像检索新的趋势。本文的主要工作有:(1)设计并实现了基于内容图像检索原型系统,该系统是进行优秀算法测试和检索效果评估的科研平台。这个系统分为图像索引和图像检索两大部分:图像索引子系统采用离线方式提取图像库中所有候选图像特征,保存在本地;图像检索子系统在线执行,提取示例图像特征并进行相似性匹配,返回检索结果。(2)深入分析了一种基于区域模糊集图像检索算法,在该算法的基础上,结合“面积优先”和“背景优先”,提出了颜色、纹理和形状特征间新的权重分配方案,并对形状特征实行特殊的匹配方案。基于区域模糊集图像检索算法采用动态k-均值聚类分析算法将图像分割成一个个区域,基于模糊集对每个区域提取模糊特征,定义了新的特征匹配方法对示例图像与候选图像进行相似性度量。实验结果表明,改进后的算法更符合人们对图像视觉特征的理解,具有良好的检索效率和检索结果。更多还原

【Abstract】 With the development of the technology of Internet and computer vision, the volume of digital image archives is growing as an unbelievable speed. The very large repository of digital information raises challenging problems in retrieval and various other information manipulation tasks. Content-Based Image Retrieval (CBIR) is aimed at efficient retrieval of relevant images from large image database based on automatically derived imagery features. These features are typically extracted from shape, texture, or color properties of query image and image in the database. The relevance between a query image and any target image is ranked according to a similarity measure computed from the features.A region-based image retrieval system segments images into regions or objects and retrieves images based on the similarity between regions. This is consistent to human recognition. In addition, applying fuzzy processing techniques to CBIR has been extensively investigated in the literature. To improve the robustness of a region-based image retrieval system against segmentation related uncertainties,“A Region-based Fuzzy Feature Approach to Content-based Image Retrieval”is selected as the theme of this paper.In the paper, we develop a prototype system started with the accuracy and real time of the content-based image retrieval. We introduce a region-based image retrieval algorithm based on fuzzy logic theory. Every core technology applied in the algorithm is analyzed in detail, including image segmentation, fuzzy feature extracted and fuzzy feature match scheme. We propose a new weight scheme combined area percentage scheme and border favored scheme to provide an overall image-to-image and intuitive similarity quantification. In addition, we do some special operations to shape feature match since we cannot have excellent shape representation.更多还原