研究海空复杂背景下红外图像小目标检测问题,针对海空背景下远距离小目标总是出现在水天线附近的特点,通过小波多尺度分析边缘检测算法检测水天线,确定出目标潜在区域,再进行精细目标搜索,从而提出了一种由粗到精的小目标检测算法.实验结果表明,对一般噪声条件下得到的红外目标图像,该方法能准确地检测、定位小目标,单帧检测准确率达96%以上并具有单帧检测率高、实现速度快的特点.

The small target detection in infrared imagery with complex background of sea and sky is studied. Taking into account that small target of infrared image guidance like naval vessel is always nearby the sea-sky-line, wavelet multiresolution edge detection algorithm is adopted to detect the sea-sky-line. The potential target area can be decided subsequently, then the target is found by careful search in this area. Therefore, a new algorithm for small target detection from coarse scale to elaborate scale is proposed. Experimental results indicate that the method can detect and locate small targets precisely in infrared image obtained in conventional noise conditions. The correct detection ratio of single image is more than 96%, and this method has high detection ratio from a single image and high speed.