摘要
本文提出了一种基于灰度形态学累加和SUSAN算法的红外弱小运动目标检测方法。首先利用Butterworth滤波器对原始红外图像进行高通滤波,得到包含少许噪声点和目标点的处理图像;然后,通过基于灰度形态学的多帧累加的方式进一步提高图像的信噪比;最后利用SUSAN检测算子对多帧累加过的红外图像进行分割并将真实目标检测出来。为了提高小目标检测的实时性,给出了基于FPGA+DSP的硬件实现结构。实验表明,该方法能够较好地消除背景和抑制噪声,从而准确有效地检测红外运动弱小目标。
In this paper, a method for detecting dim small moving infrared target based on gray morphological accumulation and SUSAN algorithm is proposed. First, a Butterworth high pass filter is used to process the original infrared image, thus an image containing some noise and target points is obtained. Then, the ratio of signal to noise of the image is further enhanced through the multi-frame accumulation in gray morphology. Finally, the SUNSAN detection algorithm is used to process the image which is multi-frame accumulated and detect the true target. In order to improve the real time performance of small target detection, a hardware structure based on FPGA and DSP is given. The experimental result shows that this method can eliminate background and depress noise effectively and hence detect the moving small infrared target accurately and effectively.
出处
《红外》
CAS
2008年第12期27-32,共6页
Infrared
基金
中国一航集团航空科学基金重点实验室类资助项目(20070112001)
