摘要
本论文主要针对现在日益增长的多媒体信息检索需求,以及目前国内外现有的基于区域的图像检索算法的局限性,深入地探讨了基于区域的图像检索的通用技术问题,提出了解决方案,着重于新的图像特征描述方法,相似度检测、特征提取与快速检索算法等方面的研究,构建了一个基于区域的图像检索通用模型,通过大量实验,验证了方案的合理性。
This thesis mainly for multimedia information retrieval is now growing demand at home and abroad, and the existing limitations of the retrieval algorithm based on region of the image, deeply discusses the general technical problem in region based image retrieval, propose solutions, focuses on the new features of image description methods, research on similarity detection, feature extraction and algorithm for fast retrieval aspects, constructs a general model of region based image retrieval, through lots of experiments, verified the rationality of plan.
出处
《科技视界》
2015年第18期46-46,91,共2页
Science & Technology Vision
基金
吉教科合字【2014】第641号
关键词
基于区域的图像检索
特征提取与匹配
图像分割方法
Region based image retrieval
Feature extraction and matching
Image segmentation method
