Wu Zimu, Wang Xia
Key Laboratory of Optoelectronic Imaging Technology and System, Ministry of Education, School of Optoelectronics, Beijing Institute of Technology, Beijing 100081, China.
Sensors (Basel). 2020 Feb 7;20(3):900. doi: 10.3390/s20030900.
With medium wave infrared (MWIR) focal plane array-based (FPA) compressive imaging (CI), high-resolution images can be obtained with a low-resolution MWIR sensor. However, restricted by the size of digital micro-mirror devices (DMD), aperture interference is inevitable. According to the system model of FPA CI, aperture interference aggravates the blocky structural artifacts (BSA) in the reconstructed images, which reduces the image quality. In this paper, we propose a novel DMD mask design strategy, which can effectively suppress BSA and maximize the reconstruction efficiency. Compared with random binary codes, the storage space and computation cost can be significantly reduced. Based on the actual MWIR FPA CI system, we demonstrate the proposed DMD masks can effectively suppress the BSA in the reconstructed images. In addition, a new evaluation index, blocky root mean square error, is proposed to indicate the BSA in FPA CI.
利用基于中波红外（MWIR）焦平面阵列（FPA）的压缩成像（CI），可以用低分辨率的MWIR传感器获得高分辨率图像。然而，受数字微镜器件（DMD）尺寸的限制，孔径干扰不可避免。根据FPA CI的系统模型，孔径干扰会加剧重建图像中的块状结构伪影（BSA），从而降低图像质量。在本文中，我们提出了一种新颖的DMD掩模设计策略，该策略可以有效抑制BSA并最大化重建效率。与随机二进制编码相比，存储空间和计算成本可以显著降低。基于实际的MWIR FPA CI系统，我们证明了所提出的DMD掩模可以有效抑制重建图像中的BSA。此外，还提出了一种新的评估指标——块状均方根误差，以指示FPA CI中的BSA。
