基于视觉对比度机制的红外小目标检测

复杂背景条件下的红外小目标检测是红外预警、红外搜索与跟踪等系统的关键技术和研究热点之一。针对红外序列图像中弱小目标检测问题,提出了一种基于视觉对比度机制的红外小目标检测方法。该方法首先运用加权高斯差分方法计算出目标显著性图,接着采用模糊控制方式优化参数获取显著性区域,最后通过显著区域与周围区域对比度分析获取真实目标。从实验结果以及和其他方法对比可以看出,所提出方法具有较高的检测率和较低的虚警率,是一种有效的小目标检测方法。

Infrared small target detection based on visual contrast mechanism

Infrared small target detection under complicated backgrounds is a key technology and a hot research in the infrared early warning system,infrared search and tracking system,etc. For dim target detection in infrared images,an infrared small target detection method was proposed based on visual contrast mechanism. Firstly,the saliency maps of the targets were calculated by the weighted Gaussian difference method. Then regions of interest were obtained by adjust parameters of fuzzy control. Finally,the contrast between regions of interest and surrounding regions was analyzed to obtain the true targets. The experimental results show that the method has the higher detection rate and lower false alarm rate,and it is an efficient infrared small target detection method.