| 闫河,刘婕,杨德红,王朴,金炜.基于特征融合的粒子滤波目标跟踪新方法[J].光电子激光,2014,(10):1990~1999 |
| 基于特征融合的粒子滤波目标跟踪新方法 |
| A new particle filter object tracking algorithm based on feature fusion |
| 投稿时间：2014-06-22 |
| DOI： |
| 中文关键词: 目标跟踪 粒子滤波(PF) 颜色特征 GMLBP纹理算子 特征融合 |
| 英文关键词:object tracking particle filter (PF) color feature global median local binary pattern (GMLBP) texture operator feature fusion |
| 基金项目:国家自然科学基金(61173184)、宁波市自然科学基金(2011A610192)和重庆理工大学研究生 创新基金(YCX2013219)资助项目 (1.重庆理工大学 计算机学院,重庆 400054; 2.宁波大学 信息科学与工 程学院,浙江 宁波 315211) |
|
| 摘要点击次数: 1506 |
| 全文下载次数: 149 |
| 中文摘要: |
| 针对传统粒子滤波(PF)算法采用单一颜色特征建模 跟踪目标性能差的缺陷,提出一种颜色特征与纹理特 征相融合的PF目标跟踪新算法。首先,采用一种具有抗噪声和保护纹理边缘的全局中值二值 模式 (GMBP)纹理算子,对模板图像进行局部差绝对值处理,得到幅 值序列模板,将幅值序列模板内的中值作为模板的阈值,与模板邻域比较获得新的纹理图像 ；然后,与 具有光照不变特性的局部二值模式(LBP)纹理算子结合,形成一种(GMLBP)新的纹理描述算子 。最后,分别计算GMLBP纹理特征粒子权值和HSV颜色特征粒子权 值,并依据权值大小确定融合系数,对纹理特征粒子权值和颜色特征粒子权值进行线 性融合,再对融合后粒子权值进行归一化处理,从而得到目标位置状态的最终估计值。对比 实验结果表明, 相对于单一颜色特征的目标跟踪算法,所提算法捕捉目标位置准确且具有更低的平均跟踪误 差,其平均误差降低了近2倍。 |
| 英文摘要: |
| For improving the performance of the t raditional particle filter object tracking algorithm based on single color feature,a new particle filter object tracking algorithm based on color an d texture feature fusion is presented in this paper.At first,a new global median local binary pattern (GMLBP) texture operator is presented,which is good for noise suppression and texture edges protection.The operator gets an am plitude sequence template by using local differential absolute value for the template image,then center pixels are replaced by median values of the amplitude sequence template,and the new texture image is obtained by being compa red with the template neighborhood. Secondly,combining with illumination invariance local binary pattern (LBP),anot her new global median local binary pattern (GMLBP) texture description operator is proposed.The fused coefficient is determined according to the different particle weights based on the GMLBP texture character istic and HSV color characteristics, respectively.The final estimation of target position is proposed via normalizing the fused particle weights.Compared with target tracking algorithm based on single color feature,experimental resul ts show that the proposed algorithm captures the target location accurately and has a lower average tracking error, and the average error is reduced by nearly 2times. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|