含噪光学遥感图像海面弱小舰船目标检测

发布时间：2018-06-23 14:57

本文选题：成像系统 + 显著性检测　；参考：《光学学报》2017年10期

【摘要】：海面舰船目标的检测与识别对于海面监测与目标打击具有重要意义,弱小舰船目标由于缺少纹理信息且易受到海面阴影、噪声等因素的影响,使得目前常用的检测方法效果较差。基于通道分离与负值扩展对比敏感函数提出了的海面弱小舰船目标检测方法。该方法首先构建像素强度通道与噪声-边缘通道的多分辨图像尺度金字塔;之后,构建不同尺度空间下的负值扩展对比敏感度函数,调制对应各位置的权重;最后,利用各空间尺度系数加权获得两通道视觉显著性图像,通过通道差分处理实现了含噪图像中弱小舰船目标的快速检测。实验结果表明:与其他5种算法相比较,提出的方法具有较高的检测准确率(97.30%)、召回率(84.71%)及综合评价指标(94.49%),同时具备较强的抗噪声能力,适用于含噪海面光学遥感图像中弱小舰船的检测。
[Abstract]:The detection and recognition of ship targets on the sea surface is of great significance for sea surface monitoring and target attack. The small and weak ship targets are vulnerable to sea surface shadow and noise because of the lack of texture information. So that the current commonly used detection methods are poor. Based on channel separation and negative value expansion contrast sensitivity function, a new method for small and weak ship target detection on sea surface is proposed. Firstly, the multi-resolution image scale pyramid of pixel intensity channel and noise-edge channel is constructed. Then, the negative value spread contrast sensitivity function in different scale space is constructed to modulate the weights corresponding to each position. Two channel visual saliency images are obtained by weighted spatial scale coefficients, and the fast detection of small and weak ship targets in noisy images is realized by channel differential processing. The experimental results show that the proposed method has higher detection accuracy (97.30%), recall rate (84.71%) and comprehensive evaluation index (94.49%) compared with other five algorithms. It is suitable for the detection of small and weak ships in optical remote sensing images of noisy sea surface.
【作者单位】：中国科学院长春光学精密机械与物理研究所小卫星技术国家地方联合工程研究中心;中国科学院大学;
【基金】：科技部重点专项(2016YFB0501202) 吉林省科技发展计划项目(20170101164JC)
【分类号】：TP751

【相似文献】