信息数据缺失下运动目标速度偏差检测仿真

摘要

对运动目标速度偏差的检测,能够更为精准预测目标运动轨迹.信息数据缺失下的运动目标速度偏差的检测,需要根据运动图像阴影区域的差分信息,对连续两帧中运动目标进行匹配.传统方法首先获取运动目标速度偏差检测的空间约束条件,降低运动目标模型的冗余计算,但忽略了对连续两帧中运动目标的匹配,导致检测精度偏低.提出基于水平集的信息数据缺失下运动目标速度偏差检测方法.通过对视频序列进行对称差分获取运动图像信息,将其融合至水平集演化函数中,得到运动目标的位置信息,调整运动图像水平集演化函数,对其进行二次分割可准确分离获得多个运动目标,对相同时域视频序列连续两帧中的运动目标进行匹配,计算出运动目标在空间域内的准确位置,实现运动目标速度偏差的检测.实验结果表明,所提方法对目标速度偏差的检测精度较高.%This research puts forward a detection method for velocity deviation of moving target under information data missing.Information of moving image was acquired via carrying out symmetric difference for video sequence,and it was fused into evolution function of the level set.Then location information of the moving target was obtained and the evolution function was adjusted.The research carried out re-segmentation for the function to separate and obtain multiple moving targets accurately and matched the moving target in consecutive frames of video sequence of the same time domain to work out accurate location of the moving target in spatial domain.Experimental results show that the method has higher detection precision of velocity deviation of target.