红外与毫米波学报, 2018, 37(3): 296, 网络出版: 2018-03-01

基于低秩矩阵近似的低噪宽幅热红外成像技术

Denoised and wide swath thermal imaging technology based on low rank matrix approximation
作者单位

1中国科学院上海技术物理研究所 空间主动光电技术重点实验室, 上海 200083

2中国科学院大学, 北京 100049

摘要
为解决强背景弱信号场景下热红外成像系统噪声制约图像信噪比的问题, 提出了一种基于低秩矩阵近似理论的低噪宽幅热红外成像技术.利用面阵摆扫方式实现宽幅扫描成像并构建严格的观测矩阵, 通过加权核范数最小化方法求解去噪的低秩矩阵形式.试验证明该技术具有较高的峰值信噪比与降噪鲁棒性, 在宽幅成像的同时也提高了探测灵敏度.研究成果在红外弱目标识别、广域侦查等领域具有一定应用价值.
Abstract
A denoised and wide swath thermal imaging technology based on low rank matrix approximation (LRMA) was proposed to solve the problem of low signal-to-noise ratio (SNR) of thermal imaging system noise in strong background and weak signal scene. This technology utilized array whisk broom mode to realize wide swath imaging and construct strict observations. The denoising low-rank matrix of observations was solved via weighted nuclear norm minimization (WNNM). Experiments showed that the technology had high peak signal to noise ratio (PSNR) and denoising robustness. Both of imaging width and detection sensitivity were improved. The technology has certain application value in the field of weak target recognition and large area investigation.
00 11