基于背景抑制与特征融合的红外微小目标检测

Small Infrared Target Detection Based on Background Suppression and Features Fusion

为了解决多特征融合微小目标检测算法复杂、受到假设限制等问题,提出了一种背景抑制与特征融合相结合的海天背景红外微小目标单帧检测算法,算法采用高通滤波抑制背景,利用灰度变换强化目标特征,特征融合时不涉及像素行均值问题,克服了海天线水平的假设;采用计算量小的局部灰度最大值、局部对比度均值反差两个特征进行加权信息融合,形成特征图,检测出微小目标.实验结果表明,该算法不论在实时性、实用性还是有效性方面都取得了满意效果.

To solve the problems of complexity and under hypothesis of the small infrared target detection algorithm based on multi-feature fusion,a novel single frame small infrared target detection algorithm of sea-sky background by combining the background suppression and feature fusion together is proposed,which adopted high pass filter to suppress background and gray-scale transformation to reinforce target features,had no concerns about row mean of pixels when feature fusion and overcomed the horizontal Sea-Sky-line hypothesis.Also,two small calculating amount features such as Local Maximum Gray Level and Local Contrast Mean Difference are adopted to fuse weighting information,then small target can be detected out from the formed feature map.The results of the experiment indicate that the proposed algorithm has achieved satisfying effects from such perspectives as timelineness,practicality and validity.