颜色与纹理自适应融合的粒子滤波跟踪算法
发布时间：2019-03-22 06:15
【摘要】：针对传统粒子滤波跟踪算法中的目标模型单一、粒子退化等问题,文章提出了一种基于颜色特征与纹理特征自适应融合的粒子滤波跟踪算法,该方法选取颜色特征和纹理特征作为目标的视觉描述子,然后将2种特征的后验概率进行融合,并对目标模板进行自适应更新,进而估计出目标的状态。实验结果表明,该方法能够较好地处理背景光照变化明显、目标物体与背景颜色相近、遮挡、局部形变等干扰因素的影响,准确跟踪及定位运动目标。
[Abstract]:Aiming at the problems of single target model and particle degradation in the traditional particle filter tracking algorithm, a particle filter tracking algorithm based on adaptive fusion of color features and texture features is proposed in this paper. The color feature and texture feature are selected as the visual descriptors of the target, then the posterior probability of the two features is fused, and the target template is updated adaptively to estimate the state of the target. The experimental results show that this method can deal with the influence of disturbance factors such as background color, occlusion, local deformation and so on, and accurately track and locate moving objects.
【作者单位】： 安徽百诚慧通科技有限公司;合肥工业大学计算机与信息学院;
【基金】：安徽省科技攻关计划资助项目(1401b042009) 安徽省高等学校自然科学研究资助项目(KJ2014ZD27)
【分类号】：TN713;TP391.41
本文编号：2445326
[Abstract]:Aiming at the problems of single target model and particle degradation in the traditional particle filter tracking algorithm, a particle filter tracking algorithm based on adaptive fusion of color features and texture features is proposed in this paper. The color feature and texture feature are selected as the visual descriptors of the target, then the posterior probability of the two features is fused, and the target template is updated adaptively to estimate the state of the target. The experimental results show that this method can deal with the influence of disturbance factors such as background color, occlusion, local deformation and so on, and accurately track and locate moving objects.
【作者单位】： 安徽百诚慧通科技有限公司;合肥工业大学计算机与信息学院;
【基金】：安徽省科技攻关计划资助项目(1401b042009) 安徽省高等学校自然科学研究资助项目(KJ2014ZD27)
【分类号】：TN713;TP391.41
【相似文献】
相关期刊论文 前4条
1 邓文坛;张三同;余纯;;一种改进的粒子滤波跟踪算法的研究[J];自动化技术与应用;2008年03期
2 王爱侠;李晶皎;陆振林;王骄;;基于FPGA的粒子滤波跟踪系统的设计与实现[J];小型微型计算机系统;2013年03期
3 张晓伟;刘弘;;一种目标优化算法改进的粒子滤波跟踪方法[J];小型微型计算机系统;2014年06期
4 李春鑫;;基于变换观测模型的粒子滤波跟踪算法[J];光电技术应用;2011年03期
，本文编号：2445326
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2445326.html
