摘要
提出了一种基于分块二值化思想的图像检索方法,利用二值化阈值将图像分割为目标和背景,在此基础上进行分块,在每一块内分别提取颜色特征后再进行相似度匹配,并在检索中突出图像目标的重要性。实验证明,这种方法能够更有针对性地检索图像,能够在保证查全率的同时较大的提高查准率。
A Method of image retrieval based on partitioned binary image is presented. The Method makes use of binary threshold to divide image into foreground and background, and extracts color feature and compares them separately in every part. In the process of retrieval, the feature of foreground will be valued. Experimental results prove that the method could retrieve images more purposefully, and precision and recall have been both improved, especially precision.
出处
《微计算机应用》
2008年第3期47-51,共5页
Microcomputer Applications
关键词
二值化
分块
图像检索
binarization, partition, image retrieval
