Study On Modeling And Evaluation Of Coded Mask Infrared Imaging Systems

Infrared imaging resolutions are limited by incident wavelength, lens materials and sensitivity of detectors, while the UAV and satellite reconnaissance are in urgent need of high resolution images. Many agencies have long advocated improving the resolutions of conventional infrared imaging systems. Recently, light field cameras and coded mask imaging systems have been successfully developed based on the compressive sensing, which have introduced digital imaging processing techniques into optical systems' designing and reinforcement and given another way to acquire high resolution images beyond limitations.In this thesis, masks' effects to the point spread function are discussed based on analysis of some typical systems for coded mask imaging. Here describes a modeling method for infrared imaging system with coded mask placed in front of lenses. As imaging result for plane waves is decided by the diffraction pattern of mask and the imperfection of the lenses, the whole system is modeled as mask with ideally focused lenses and practical lenses with imperfection. A set of criteria are also presented to assess the point spread function variations to cell size, coding matrix and periodicity, which help analyzing the system's ability of light gathering and sensibility to source position. Numerical experiments show that the model matches practical measurement. The Dammann grating pattern has a balance satisfaction in both direct imaging and practical measurement.The model we build gives a definite expression for coded mask imaging system and helps further research on simulation platforms. The criteria results of different masks can be used as references to guide the design and optimization for mask pattern.