基于图像区域Lyapunov指数的海面舰船目标检测

Detection of ship targets on the sea surface based on Lyapunov exponents of image block

为了检测海面背景中的舰船目标,分析了目标存在时背景信号混沌特征的变化,提出了一种基于图像区域Lyapunov指数的目标检测新方法.新方法定义了图像灰度距离的概念,基于改进的Wolf方法将一维信号Lyapunov指数提取方法扩展到图像信号,利用图像区域最大灰度距离Lyapunov指数的变化检测淹没在混沌背景信号中的目标信号.实验结果表明海面背景图像信号具有一定的混沌特征,利用新方法能有效检测出海面背景下的舰船目标,检测结果优于基于统计分析的方法.

The change of background＇s chaos feature when there are moving targets was analysed, and a new algorithm based on Lyapunov exponents of image block for moving targets detection on the sea surface, which is used for ship targets detection on the sea surface, was proposed. The new method defines the term of image gray level distance, extends the calculation method of Lyapunov exponents from single dimension signal to image signal based on the improved Wolf method, uses the change of the largest gray level distance Lyapunov exponents to detect ship targets submerged in chaos background. Experimental results show that the sea surface background has chaos feature. The proposed algorithm can detect ship targets effectively and can get better result than statistical analysis.