复杂海天背景下红外舰船目标的自动检测方法

Automatic detection method of IR warship target in the complex sea-sky background

近年来,舰船目标的检测与识别技术得到了人们越来越多的重视。研究了复杂海天背景下红外舰船目标的自动检测问题。首先运用小波变换对图像进行预处理,提高了图像的信噪比,从而抑制了噪声和背景杂波。然后针对红外舰船目标总是出现在海天线附近的问题,采用梯度强度均值分割的算法检测了海天线,确定了目标的潜在区域。最后,提出了一种基于行列均值的自适应阈值分割算法,结合使用红外舰船目标的几何特征量,采用特征匹配方法提取出舰船目标,达到了自动检测的目的。实验结果表明:该方法能有效地自动检测复杂海面背景中的红外舰船目标。

In recant years,the warship target′s detection and recognition has attracted more attentions. There have been lots of valuable studies and papers about the warship target. Due to the importance of automatic warship target detection,the automatic detection issue of infrared warship targets in the complex background of sea-sky was studied. At first,in order to improve the image signal-to-noise ratio and restrain the noise and background clutter,an image preprocessing method based on wavelet transformation was adopted so that the target can be easily detected. Then considering the infrared ship targets are always located nearby the sea-sky-line,the mean value of grads intensity was adopted to detect the sea-sky-line so that the potential target area can be decided subsequently. Finally,an adaptive threshold segmentation algorithm based on the mean of image rows and columns was proposed. After image segmentation,ship target was extracted by using the characteristic matching method combined with the geometric characteristics of infrared warship target. According to the above steps,the issue of automatic detection of infrared warship was finally solved. The experimental results show that this method is successful and can effectively detect the infrared warship target in complex background of sea and sky.