摘要
基于内容的图像检索是数字图像处理的一个重要研究方向,有效地提取图像内容特征是其中的一个关键问题。提出利用最大相关最小距离将图像的纹理特征、高斯密度特征与人脸检测相结合的算法进行图像检索。在建立10 000幅图像库的基础上验证了算法的可行性,实验结果表明算法能够准确、高效地检索出目标图像;相对于单一特征的检索,算法有效地提高了图像检索的精度和速度。
Content-based image retrieval is an important research direction of digital image processing, and effective image extraction feature is one of the key issues. A new image retrieval method by using max correlation rain distance to combine together texture features, Gaussian density characteristics and face detection of images for image retrieval is presented. The establishment of 10 000 images to prove the feasibility of the algorithm, the experimental results show that algorithm can be accurately and efficiently retrieve target image. Compared to a single retrieval features algorithm that it improves effective image retrieval accuracy and speed.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第17期4507-4510,共4页
Computer Engineering and Design
基金
重庆市科委自然科学基金项目(CSTC-2006BB2309)
关键词
基于内容的图像检索
高斯密度特征
纹理特征
人脸检测
最大相关最小距离
content-based image retrieval
Gaussian density feature
texture feature
face detection
max correlation rain distance
