A Focal Plane Coding Method for Large Field of View and Variational Resolution Imaging

摘要

Remote sensing imagery is an important method for obtaining battlefield information,which requires variable field of view (FOV) and resolution for different reconnaissance mission. A focal plane coding method for variational resolution imaging with different FOV based on compressed sensing (CS) is presented in this paper. Combined with the control of several lenses,a programmable coding mask called Digital Micro-minor Device (DMD) is placed on the focal plane of the optical system,which makes what the detector measures is the compression sample of the observed image,enables variational resolution imaging possible. On the basis of this framework,we design a programmable DMD and a corresponding lens array according to imaging modes. For full scene observation,differnent lens sub-array observes different sub-region and nxn micro-mirrors are regarded as a pixel,making it with a large FOV and low resolution. For region of interest (ROI) observation,all the lenses observe the same region and every micro-mirror is thought to be a single pixel,making it with high resolution. The object is sampled sparsely and repeately using multiplexing at one snapshot. A smart reconstruction algorithm is adopted for image recovery. Numerical simulation shows that the proposed method is feasible,which enables imaging modes to swith between large FOV and high resolution flexibility and ensures the imaging quality.