针对海洋背景目标检测,提出了一种基于偏振信息融合的海洋背景目标检测方法。该方法根据目标与水体偏振信息的差异性以及偏振图像能很好弱化 背景噪音的特点,设计了一组图像融合规则,并通过该组规则来融合强度、偏振度和偏振角等各分量图像。最后,对融合结果采用基于轮廓层次树的目标 检测方法进行测试。实验结果表明,融合后的图像能很好地抑制海面波纹噪音,同时增强了目标的边缘轮廓和纹理细节信息,以及提高了目标与背景 的对比度,使目标更易于被检测出来。

According to target detection at sea, a method for target detection of ocean background based on polarization imagery fusion is presented. This method make use of the polarization information’s difference between the target and water, and good restrain of background noise. A group of image fusion rules are designed, and intensity of images, degree and phase of polarization images are fused by this rules. Finally, the results of fusion are tested by using the target detection method based on contour hierarchy tree. The experimental results show that the method can strengthen the target’s edge and texture features, enhance the contrast of target and background, which makes the target easier to be detected.