基于二维熵的空中小目标跟踪研究

二维熵是图像分割技术中的一种重要方法。结合空中小目标的特点将二维熵阈值算法应用于空中小目标图像序列中,并求取质心进行跟踪。重新定义了传统二维熵中二维直方图的双轴,并加以讨论。同时用泰勒一阶展开求取公式的近似表达,消除原始算法中大量的对数计算,加快了运行的速度。仿真实验表明,该方法可以有效的跟踪空中运动目标,在工程实际中具有很好的应用价值。

Research of small aerial target tracking based on two-dimension entropy

Two-dimension entropy is one of the important methods in image segmentation. Considering features of the small aerial target,this paper uses two-dimension entropy in small aerial target motion sequences,calculates the centroid for tracking, redefines axis of coordinates in 2D histogram which is used in traditional two-dimension entropy. Then dis- cusses it. Taking the first Order Taylor series expansion as approximate calculation can get rid of numerous logarithmic computations and speed up on Digital Signal Processor,which proved feasible on theory. The results of simulation experi- ment confirmed that this method could be effective to track the small aerial target,practical and worthy of using abroad.