摘要
针对强噪声背景下合成孔径雷达图像中舰船检测困难的问题,提出了基于剪切波(Shearlet)变换舰船检测方法。首先利用Shearlet变换分解原图像;然后根据Shearlet高频系数在目标区域和背景区域具有不同的表现性质,将多方向多尺度的Shearlet系数进行融合,实现了噪声抑制和舰船目标增强;最后采用阈值方法分割出舰船目标。实测SAR图像数据的实验表明,所提出的检测方法在强噪声背景下,相对于传统恒虚警率方法和基于小波加强的方法,能够达到较高的检测概率和较低的虚警率。
To solve the difficulty in ship detection for SAR image in the case of the strong sea clutter background,a ship detection method based on Shearlet transform is proposed. At first, SAR image is composited by Shearlet transform; then Shearlet coefficients at muhiscale and multi-direction are fused to enhance the ship target and reduce the noise because the Shearlet coefficients of target and background at different scale and different direction have different property. At last,the threshold detector is used to segment the ship targets from the fused Shearlet coefficients image. Through the experiments on real SAR image, the proposed method can reach higher detection rate and lower false alarm rate in the case of the strong sea clutter background relative to the traditional CFAR detector and wavelet enhancement method.
出处
《电子测量技术》
2014年第6期54-58,62,共6页
Electronic Measurement Technology
基金
国家自然科学基金(61101201)资助项目
