摘要
提出一种感兴趣区域的图像检索方法,该方法首先采用遗传FCM算法对图像进行分割,然后提取分割后的区域特征进行检索;为了缩小低层特征和高层语义之间的语义鸿沟,最后提出一种基于神经网络的相关反馈方法,实验表明,该方法具有较好的检索性能,系统具有较高的查全率和查准率.
An image retrieval method based on region of interest is proposed. First, genetic fuzzy c-means （FCM） is used for image segment, and then the feature of region of interest is extracted. In order to reduce the semantic gap between the lowlevel visual features and the high-level semantic content, a relevance feedback method based on neural network is proposed. Experimental results show that the proposed image retrieval algorithm has a better retrieval performance.
出处
《河南师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2009年第1期33-36,39,共5页
Journal of Henan Normal University(Natural Science Edition)
基金
国家863基金(2005AA414010)
浙江省自然科学基金(M603034)
