基于压缩感知和红外玫瑰线扫描的红外图像重构（英文）

发布时间：2018-11-05 18:43

【摘要】：红外亚成像制导技术是由点源探测技术到成像制导技术的一种过渡,由单元探测器和光机扫描装置组成.红外玫瑰线扫描亚成像系统是亚成像制导中的一种,红外玫瑰线扫描亚成像系统按照特定的图案采集视场中的部分数据并得到一幅含有目标位置信息的亚图像.受单像素相机的启发,主要研究红外玫瑰线扫描亚成像系统中的压缩成像.压缩感知可以在更少的采样数据条件下重构红外图像,其应用到红外亚成像制导系统中一个关键的问题就是观测矩阵的构造.关于随机观测矩阵的研究已经比较广泛,但随机矩阵很难实现.本文提出了一种简单的适用于红外玫瑰线扫描亚成像系统的确定性观测矩阵.此外还提出了一种快速有效的恢复算法,称为优化子空间追踪算法.仿真结果显示构造的观测矩阵能够压缩和重构红外图像,且重构效果优于随机高斯观测矩阵和随机伯努利观测矩阵,提出的恢复算法也具有较好的表现.
[Abstract]:Infrared subimaging guidance technology is a transition from point source detection technology to imaging guidance technology. It is composed of unit detector and optical scanner. The infrared rose line scanning sub-imaging system is one of the sub-imaging guidance systems. The infrared rose line scanning sub-imaging system collects some data from the field of view according to a specific pattern and obtains a sub-image containing the target position information. Inspired by single pixel camera, compression imaging in infrared rose line scanning sub-imaging system is studied. Compressed sensing can reconstruct infrared images with less sampled data. One of the key problems in infrared subimaging guidance systems is the construction of observation matrices. The research of random observation matrix has been widely studied, but it is difficult to realize the random matrix. In this paper, a simple deterministic observation matrix for infrared rose line scanning subimaging system is presented. In addition, a fast and effective recovery algorithm, called optimal subspace tracking algorithm, is proposed. Simulation results show that the constructed observation matrix can compress and reconstruct infrared images, and the reconstruction effect is better than that of random Gao Si observation matrix and random Bernoulli observation matrix.
【作者单位】：哈尔滨工程大学信息与通信工程学院;
【基金】：Supported by National Natural Science Foundation of China(61571146) Natural Science Foundation of Heilongjiang Province of China(F201407)
【分类号】：TN911.73

【相似文献】