基于小波变换的海面目标自动检测

Automatic Sea Target Detection Based on Wavelet Transform

利用小波分析优良的局部化能力和方向特性,提出了一种基于小波变换的海面目标自动检测算法,该算法在小波分解的垂直分量中进行目标检测,经过互能量交叉和海杂波抑制后,通过最大能量判决检测目标位置,并利用双窗口相似性度量函数标定目标大小。实验结果表明,该算法可以在单帧图像中检出目标,并且对存在云层、鱼鳞光等干扰的复杂海空背景也能取得较好的检测效果。

An effective target detection algorithm based on wavelet transform was proposed by nice lo- calization ability and directional characteristic of wavelet multi-resolution analysis. The target detection was conducted by the algorithm in vertical component of wavelet decomposition. After the mutual energy combination and the sea clutter suppression, the target position was detected by the maximum energy judgement, the target size was calibrated by the double window similarity measurement function. The experimental result shows that the target can be detected by the algorithm in single frame image, the better detected effect can be obtained under the complicated background of sea and sky with the disturbance of cloud and fish scale light.