摘要
前视红外(FLIR)小目标图像背景比较复杂,给目标检测和识别造成较大困难。针对图像背景中存在杂波,提出一种基于形态重构的背景抑制技术,分析了获取标记图像时结构算子的选取方法。对AMCOM数据集上的实验表明,提出的算法对于提高图像的信噪比是很有效的。
The background is very complex in forward-looking infrared （FLIR） image, which result in very difficult to detect and track small target. According to the clutter in the background, we propose a background reduction algorithm using morphological reconstruction operators, and analysis the selection of structuring element to get the marker image. Experimental results, obtained with the AMCOM FLIR data sets, demonstrate the effectiveness of the proposed method on improving the signal-to-noise （SNR）.
出处
《红外技术》
CSCD
北大核心
2008年第12期713-716,共4页
Infrared Technology
关键词
形态重构
结构算子
背景抑制
morphological reconstruction operator： structuring element
background reduction
