基于H∞滤波器的红外小目标运动预测和跟踪方法

根据H∞滤波理论，提出了基于H∞滤波预测技术的红外图像小目标运动预测和跟踪方法，为了降低模型阶数以加快图像处理速度，滤波过程被分解在水平和垂直两个方向上分别进行，即将滤波器分解为x、y方向上两个子滤波器．H∞滤波器是一种基于最优化最坏情况下的滤波技术，它对噪声源的要求不高，能较好地应用在复杂场景下的目标运动预测与跟踪．实验证实了在噪声源不能确定或是未知的情况下，它比Kalman滤波器对红外小目标的运动预测更加准确并具有更好抗扰动性．

APPROACH OF INFRARED SMALL TARGET MOTION PREDICTION AND TRACKING BASED ON H∞ FILTER

According to the theory of H_∞ filter,an approach of small target motion prediction and tracking in infrared images was presented based on H_∞filter technology.In order to reduce the rank of the model for speeding up the velocity of image process,the filter process was analysed along horizontal orientation and vertical orientation,that is,the filter was decomposed into two filters along x and y orientation.H_∞filter is a filter technology based on the optimization of worst condition which doesn't require too much about the noise sources and it can be applied to target motion prediction and tracking in complex and clutter background.The experiments demonstrate that H_∞ filter is more exact for IR small target motion prediction and resistible to the disturbance than Kalman filter in the system which has the uncertain or unknown noise sources.