Encoded Optical Compressive Imaging Architectures And Reconstruction Algorithms

Traditionally, optical sensors are designed to collect the most directly interpretableand intuitive measurements possible. With the emerging framework of compressedsensing (CS), however, indicate that substantial performance gains may be possible vialess direct measurements combined with computational methods, besides, whenfeasible selection of the type of measurement systems may dramatically improve theability to perform image reconstruction. In particular, by designing optical sensors todeliberately collect"incoherent"measurements of a scene according to CS theory, itcan use sophisticated computational methods to infer more information about structureand content of original images. These ideas are particularly relevant to imagingapplications in which it is often useful to keep the FPA of an optical system relativelysmall. However, applying this theory to practical imaging systems is challenging in theface of several measurement system constraints, it is diffcult to actually measure therandom projections described in the literature, and therefore, innovative andsophisticated imaging systems should be ingenious designed. In this work, it introducea compressive optical imaging techniques that combines coded aperture for theextraction of high-resolution images from relatively small focal plane arrays.Furthermore, it explore a different physically realizable optical systems that combines4f imaging system for acquiring such measurements based on CS principles andcombined with sophisticated sparsity optimization algorithms can significantlyincrease image quality and resolution.