基于多小波分析和数据拟合的小目标跟踪方法

针对红外序列图像弱小目标的跟踪问题,提出了将多小波分析与数据拟合结合起来的小目标跟踪方法。利用对多小波分析进行目标奇异性的检测,得到了可能的点目标利用数据拟合方法在连续多帧内对每组可能点数据进行拟合,根据运动目标轨迹特性判断出真实目标,进而进行跟踪。仿真试验表明,该方法能有效地实现低信噪比下的小目标稳定跟踪.

Tracking dim moving target based on multi-wavelet analysis and data fitting

Aiming at the dim target tracking in the infrared sequence image,a dim target tracking method based on both the multi-wavelet analysis and the data fitting is brought forward.The singularity detection of the target using the multi-wavelet analysis is carried out,and the likely dot target is got.The likely data of every group in a seriate frame are fitted.According to the tracking character of moving target,the real target is determined,then it is tracked further.The simulation experiment shows that the method can realize stable tracking of the dim target efficiently in the condition of low SNR.