红外图像序列中运动面目标的分割算法研究

利用红外成像实现自动目标检测,这在现代防御技术中具有重要的地位。本论文主要针对红外预警系统的目标探测补盲阶段,深入研究了现有的比较成熟的各种运动目标分割算法,用软件仿真了各算法处理图像的效果,并总结归纳了各种算法的优缺点,提出了在差分图像的基础上,采用二维最大熵法对图像进行分割以更好地提取区域目标的观点。经实验证明,该算法效果良好。

Study of Algorithm for Detecting Regional Moving Target in Infrared Image Sequences

It is very important to use infrared imaging for target automatic detection in the modern defense technology. In this paper, all kinds of common segmentation algorithms of moving targets which can be used in the shadow compensation phase of target detection of infrared warning systems are further studied. The effectiveness of each algorithm in processing images is simulated by using a software. The advantages and disadvantages of different algarithms are summarized. A 2D maximum entropy approach based on the subtraction between two consecutive frames is proposed. The experimental results have shown that the proposed approach is effective for the detection of regional moving targets in infrared image sequences.