| 基于ICA子空间的目标运动变化检测方法 |
| |
| 引用本文： | 张桂林,张波. 基于ICA子空间的目标运动变化检测方法[J]. 华中科技大学学报(自然科学版), 2005, 33(12): 79-81 |
| |
| 作者姓名： | 张桂林 张波 |
| |
| 作者单位： | 华中科技大学,图像识别与人工智能研究所,湖北,武汉,430074;华中科技大学,图像识别与人工智能研究所,湖北,武汉,430074 |
| |
| 基金项目： | 国防科技重点基金资助项目. |
| |
| 摘 要： | 针对实时场景下复杂背景动态变化的特点，在子空间图像差分模型的基础上，提出了一种基于ICA子空间的目标运动变化检测方法，该算法能够检测出两帧图像之间目标运动变化而不是背景变化部分，克服了目前大多数基于简单差分模型的变化检测算法对图像配准精度要求很高的问题．实验结果表明，这种方法能够准确地提取出实时场景中目标的运动变化区域，并具有很强的鲁棒性．
|
| 关 键 词： | 图像差分模型 独立分量分析 子空间 目标运动变化检测 |
| 文章编号： | 1671-4512（2005）12-0079-03 |
| 收稿时间： | 2005-03-18 |
| 修稿时间： | 2005-03-18 |
Detection method for change of targets movement based on subspace of independent component analysis |
| |
| Zhang Guilin,Zhang Bo. Detection method for change of targets movement based on subspace of independent component analysis[J]. JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE, 2005, 33(12): 79-81 |
| |
| Authors: | Zhang Guilin Zhang Bo |
| |
| Abstract: | According to the features of dynamic changes of the complex background in real-time sceneries, a new method for the change detection of moving targets was proposed with the help of the image differencing model of subspace. In this method, only moving of the targets is detected well and the changes of back- ground are not done. Therefore, the majority of the current change detection algorithms based on simple image differencing models are not largely dependent on the precision of geometric alignment for imagery. The experiment results indicated that the method is capable of accurately detecting change areas of moving targets in real-time sceneries with strong stability. |
| |
| Keywords: | image differencing model independent component analysis(ICA) subspace targets moving change detection |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
