摘要
考虑到目前许多基于颜色直方图图像检索系统的搜索质量往往相当有限,提出一种融合分块小波直方图相似度检索和粒子群优化的新方法.该算法引入小波技术,提高了特征提取的有效性,采用分块技术扩展了图像检索性能,结合微粒群算法进行智能搜索加快了算法的执行速度.实验结果证实,该算法对图像数据库的相似度搜索是切实可行的,为大型图像数据库的智能图像检索问题提供解决方案.
Considering that the quality of the outcomes provided by color histogram-based image search was usually rather limited when it was applied to fast similarity search in the large image databases,which was based on blocking wavelet-histogram image similarity retrieval method and particle swam optimization(PSO),an innovative approach was proposed.The experimental results showed that this algorithm by combining blocking wavelet transformation and particle swarm optimization with color histogram-based image retrieval could get the higher retrieval performance and the quicker speed to the similarity search in images database.
出处
《湖北大学学报(自然科学版)》
CAS
北大核心
2010年第4期379-383,共5页
Journal of Hubei University:Natural Science
基金
湖北省教育厅重点科研项目(D20101704)资助
关键词
图像相似度检索
分块小波变换
粒子群优化
颜色直方图
image similarity retrieval
blocking wavelet transformation
particle swam optimization
color histogram