一种基于自适应背景预测的红外弱小目标检测方法

针对目前典型的基于背景预测的红外弱小目标检测方法中无法较好消除噪声的问题,提出了一种基于自适应背景预测的红外弱小目标检测方法。该方法不同于传统方法,在对红外图像进行背景预测时,不仅考虑邻域内像素点的灰度值,同时还通过自适应处理的方法将待预测像素点灰度值也考虑其中,能有效改善低信噪比下红外图像弱小目标的检测性能。通过大量的仿真数据对本文和传统方法进行了对比和分析,证明了该方法的有效性和优越性。

Detection method for dim small infrared targets based on adaptive background prediction

Tradition background prediction often introduces extra noise.Aiming at solving the problem,a novel detection method for dim small infrared targets based on adaptive background prediction is proposed.The method is different from traditional method in many ways.When predicting the background of infrared image,not only the gray values of neighborhood pixels are considered,but also the gray values of the predicted pixels are taken into account.It can improve detection performance for low SNR (signal to noise ratio) infrared images.By comparing and analyzing this method and the traditional methods,it is proven that the new method is more effective and robust.