基于仿生形象思维方法的图像检索算法的改进

PDF下载HTML

Title:

Improvement of an image retrieval algorithm based on biomimetic imaginal thinking

作者:

陈阳¹, 董肖莉², 李卫军², 张丽萍², 覃鸿²

1. 工业和信息化部 中国电子信息产业发展研究院, 北京 100846;
2. 中国科学院半导体研究所 人工神经网络实验室, 北京 100083

Author(s):

CHEN Yang¹, DONG Xiaoli², LI Weijun², ZHANG Liping², QIN Hong²

1. China Center of Information Industry Development, Beijing 100846, China;
2. Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China

关键词:

仿生形象思维; 图像检索; 颜色复杂度; 特征提取

Keywords:

biomimetic imaginal thinking; image retrieval; color complexity measure; feature extraction

分类号:

TP391.4

DOI:

10.3969/j.issn.1673-4785.201411022

文献标志码:

A

摘要:

基于仿生形象思维方法的图像检索算法主要是模仿人脑的形象思维对图像认知,从全新角度提取图像特征而设计的一种新算法。此算法把每幅图像都映射成高维空间一个点,通过计算点和点之间的判别函数得到图像之间的关系。该文利用最直接地描述图像内容的视觉特征,即颜色复杂度来提取图像特征,对基于仿生形象思维方法的图像检索算法做进一步研究与改进。实验结果表明该方法比文献[1]基于仿生形象思维方法的图像检索算法的特征提取方法效果有一定的提高。

Abstract:

In this paper, a novel image retrieval algorithm based on biomimetic imaginal thinking is used for image cognition by imitating human brain’s imaginal thinking and extracting image features from a different perspective. This algorithm maps every image onto a point in the high dimension space, deriving the relations of two images by calculating the discriminant function between the two points. The visual features that can describe image content directly, i.e. color complexity are used to extract image features, in order to improve the image retrieval algorithm based on biomimetic imaginal thinking. The experimental results showed that the performance of this algorithm is better than that of the algorithm proposed by reference [1].