基于滑窗式单帧红外弱小目标检测方法研究

Research into The Method of Detecting The Dim Infrared Target in Single Frame Image Based on Sliding Window

提出了一种基于滑窗式单帧红外弱小目标检测定位的算法:选择合适的滑窗,针对弱小目标在滑窗局部内具有较高灰度值的特征,通过自适应阈值来判别该滑窗中心位置是否存在目标。当滑窗遍历整幅图像后,就可以得到红外小目标的检测结果。在判别的过程中,构造了一种新的自适应阈值估计的方法。通过大量的仿真实验,并与当前常用的形态学滤波、巴特沃斯低通滤波等方法的比较,证明了该算法的有效性和优越性。

This paper presents a algorithm to detect and locate the dim infrared target in single frame image based on sliding window：selecting the appropriate sliding window,aiming at the characteristics that the dim targets has higer grey value in partial sliding window,distinguishes whether there are targets or not in the middle position of sliding window through adaptive threhold.After the sliding window goes through the whole diagram,the detection result of the dim infrared targets can be fetched.In the process of judgement,a new adaptive threshold estimation method is constructed.Through a great of simulation experiments,and compared with the common methods methods such as the morphological filtering,Butterworth filtering,etc.,the validity and superiority of the algorithm are proved.