基于均值漂移的水面运动目标跟踪
Water Surface Moving Target Tracking Based on Mean Shift
提出了一种自启动的均值漂移水面运动目标跟踪算法。在跟踪的起始阶段,采用统计分析技术,从场景中提取运动目标,从而建立目标的彩色直方图,并获得目标的初始位置。在后续图像帧中,基于均值漂移算法,对面目标进行跟踪。仿真实验表明,在水面复杂环境下,该算法能自启动,并稳定处理跟踪水过程中的旋转、局部遮挡等问题。
water surface moving targets tracking algorithm based on Mean Shift is proposed in this paper, which is automatically initial. At first Differential Image method base on statistics is used to segment the moving area, so the histogram of target could be built. The initial 2D position of the target is also got. Then the target can be tracked by using mean-shift tracking algorithm in the following frames. Experiments show that the algorithm can self starts, and solves the problem of rotate and partial occlusion during tracking.
朱威
电子技术应用自动化技术、自动化技术设备计算技术、计算机技术
水面目标跟踪运动目标分割均值漂移
water surface moving targets trackingtarget segmentationmean shift
朱威.基于均值漂移的水面运动目标跟踪[EB/OL].(2010-12-23)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/201012-943.点此复制
评论