检索结果-内蒙古大学图书馆

MapReduce model for efficient image retrieval: a hadoop-based framework

International Journal of Information Technology (Singapore) 2025年第2期17卷 925-939页

作者： Alrahhal, Maher Shukla, Vinod Kumar Postdoctoral Fellow School of Engineering Architecture and Interior Design Amity University Dubai Dubai United Arab Emirates Associate Professor School of Engineering Architecture and Interior Design Amity University Dubai Dubai United Arab Emirates

The rapid proliferation of images, driven by advancements in image-capturing technologies, poses significant challenges to the efficient management and retrieval of images from vast databases. Traditional Content-Based Image Retrieval (CBIR) systems struggle with scalability and complexity, resulting in suboptimal retrieval performance. To address these challenges, this paper explores the integration of CBIR with hadoop, a well-established distributed computing framework. hadoop's capability to handle large-scale data processing, utilizing its MapReduce model and hadoop distributed File System (HDFS), offers a promising solution to enhance the effectiveness and scalability of image retrieval tasks. Our solution combines the strengths of CBIR and hadoop, using our proposed Full Directional Local Neighbor Pattern (FDLNP) method for feature extraction. This method captures local patterns, color, texture, and directional information to provide a comprehensive image representation, significantly improving retrieval accuracy. We present a detailed design and implementation of this integrated system, emphasizing its two main phases: the offline phase, which constructs the feature database using FDLNP and MapReduce jobs, and the online phase, which extracts features from query images and calculates similarity distances using parallel processing. Experimental results on a hadoop cluster reveal a significant improvement in processing efficiency, particularly for large datasets, highlighting the advantages of distributed processing in managing extensive image retrieval tasks. The findings indicate that while single-node systems may be suitable for smaller datasets, hadoop clusters are preferable for larger-scale image databases due to their scalability and concurrent processing capabilities. Integrating CBIR with hadoop, enhanced by our FDLNP method, provides a powerful tool for organizing and searching images in vast databases, offering improved retrieval performance, scalabi

关键词： Big data processing CBIR CBIR with hadoop integration hadoop distributed computing framework hadoop distributed file system (HDFS) MapReduce model

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：