Compressive Imaging has been an extensively researched area inoptical imaging, object tracking,satellite applications,etc. There are many signalrecovery methods and comparative analyses of different algorithms in the presence of Gaussian noise. However, certain applications such as optical imaging at low photon intensityhavecounts of discrete events, which cannot be modelled using a Gaussian noise model. Instead, anoise model that incorporates photon statistics is needed. Researchers haveworkedon the Poisson noise modeland a different compressive sensing reconstruction was found.In this thesis, weconsidered a more general scenario of Compressive Imaging using non-classical photon states as light sources. We assumed that the Compressive Imaging system that consists of digital micromirror device (DMD), lenses, and detectors are perfect so that all noises comes from photons. Fock states and squeezed light thatpossess non-Poissonian statistics plays animportant role in Quantum Imaging. The image reconstruction was performed using several common compressive sensing signal reconstruction algorithms assuming Gaussian noise. This thesis showedthe behavior of the root mean square error (RMSE)with respectto the signal-to-noise ratio (SNR)for different photon statistics. In particular, the study showedthat all the noises perform similarly for the different algorithms. Based on the performance results for the different light sources, this researchcan be helpful in designing ageneralized Compressive Sensing model incorporating the photon statistics thatareapplicable in the field of Quantum Optics.