文件名称:tidu
介绍说明－－下载内容均来自于网络，请自行研究使用
针对mean shift 跟踪方法中存在的光照变化不稳定问题， 提出了基于梯度特征与彩色
特征相融合的mean shift 跟踪方法。首先分别提取目标的梯度特征和彩色特征，利用多尺度的相似度
计算方法进行特征的匹配，然后通过最大化相似度对目标进行跟踪。通过物体和人体等运动目标的跟
踪，验证了改进的跟踪算法在光照变化情况下的鲁棒性优于原有的算法，显著降低了跟踪位置误差。-The instability of the light changes in the mean shift tracking method, the mean shift tracking method based on gradient features and color characteristics of fusion. First extract gradient features and color features of the target, the similarity calculation method using multi-scale feature matching, and then maximize the similarity of the target track. Through the tracking of objects and people moving target, verify the robustness of the improved tracking algorithm in the case of illumination changes is better than the original algorithm significantly reduces the tracking position error.
特征相融合的mean shift 跟踪方法。首先分别提取目标的梯度特征和彩色特征，利用多尺度的相似度
计算方法进行特征的匹配，然后通过最大化相似度对目标进行跟踪。通过物体和人体等运动目标的跟
踪，验证了改进的跟踪算法在光照变化情况下的鲁棒性优于原有的算法，显著降低了跟踪位置误差。-The instability of the light changes in the mean shift tracking method, the mean shift tracking method based on gradient features and color characteristics of fusion. First extract gradient features and color features of the target, the similarity calculation method using multi-scale feature matching, and then maximize the similarity of the target track. Through the tracking of objects and people moving target, verify the robustness of the improved tracking algorithm in the case of illumination changes is better than the original algorithm significantly reduces the tracking position error.
(系统自动生成,下载前可以参看下载内容)
下载文件列表
基于梯度特征与彩色特征相融合的meanshift跟踪方法.pdf