基于多源遥感卫星的海面舰船目标检测方法

围绕光学卫星遥感图像中的海面舰船目标检测问题,考虑静止轨道遥感卫星和低轨道高分辨率遥感卫星的优缺点,提出一种基于多源遥感卫星的海面舰船目标检测方法。该方法首先通过多结构多尺度形态学滤波和自适应阈值分割对静止轨道卫星遥感图像进行目标预检测,然后将预检测结果与卫星图像的地理位置信息相结合来确定高分辨率遥感图像候选目标区域,最后提取候选目标区域中目标的主要特征,进行舰船目标判别,将判别结果反映到静止轨道遥感卫星图像上。仿真结果表明,该方法能有效检测出海洋背景遥感图像中的多个运动舰船目标,具有目标检测率高、虚警率低、时效性好等优点,为静止轨道遥感卫星对海面运动舰船目标的实时监视跟踪提供了技术支撑。

Ship target detection method based on multi-source remote sensing satellite

Aiming at the problem of detecting ship targets on sea surface from optical satellite remote sensing images,considering the advantages and disadvantages of geostationary orbit remote sensing satellite and low orbit high resolution remote sensing satellite,a method for detecting ship targets on sea surface based on multi-source remote sensing satellites is proposed.Firstly,the target pre-detection of geostationary-satellite remote sensing images is carried out by multi-structure multi-scale morphological filtering and adaptive threshold segmentation method.Secondly,the candidate target area in the high-resolution remote sensing image is settled by combining the pre-detection result and the geographic location information of the satellite image.Finally,the main features of the target in the candidate target area are extracted,and the ship target is discriminated.The discriminated result is reflected to the geostationary orbit remote sensing satellite image.The simulation results show that the proposed method,in the remote sensing image of ocean background,can detect a number of moving ship targets effectively,and it has the advantages of high target detection rate,low false alarm rate and high timeliness.In addition,this method also provides technical support for the geostationary orbit remote sensing satellite to monitor and track the moving ship targets in real-time.