摘要
相关反馈技术是提高图像检索性能的一个重要手段,本文提出了一种新的基于多类SVM的图像相关反馈检索方法,检索结果表明多类SVM方法在检索性能上具有较其他检索方法更高的检索准确性,并且其检索的密集度较传统SVM检索方法更优.
Relevance feedback is an important research field in image retrieval. A novel image relevance Feedback algorithm is proposed in this paper. The experiment indicates that this algorithm gets higher accuracy in retrieval performance than others, also has better dense performance than the retrieval algorithm based on normal SVM methods.
出处
《华中师范大学学报(自然科学版)》
CAS
CSCD
2008年第3期355-358,共4页
Journal of Central China Normal University:Natural Sciences
基金
公安部创新基金项目(2007YYCXHBST068)
关键词
多类SVM
图像检索
相关反馈
multi-class SVM
image retrieval
relevance feedback