Date of Completion


Embargo Period



3D Imaging, Low Light Imaging, Pattern Recognition, Integral Imaging, Photon-Counting, Optical Security

Major Advisor

Bahram Javidi

Associate Advisor

John Chandy

Associate Advisor

Rajeev Bansal

Field of Study

Electrical Engineering


Doctor of Philosophy

Open Access

Open Access


There has been significant research in developing novel image sensors to capture a scene in low illumination environments. Using conventional passive imaging sensors that operate in the visible spectrum, camera read-noise becomes higher than the signal information from the scene resulting in read-noise dominant images. These images suffer from poor signal-to-noise ratio making scene visualization unobtainable. Thus, imaging in photon-limited environments has been viewed as a nuisance. This dissertation will consist of three separate parts that discuss utilizing photon-counting principals for secure information storage or imaging in photon-limited environments. First, an optical security system for information storage is introduced that uses secure three-dimensional optical phase codes whose unique optical signature can be authenticated using the random-forest classifier. Next, a system for securing three-dimensional integral imaging displays using a quick-response encoded elemental image array will be discussed that combines photon-counting with traditional optical encryption methods for secure information storage. The second part of the dissertation will discuss integrated circuit authentication in photon-limited environments. More specifically, a photon-counting model will be applied to an image of an integrated circuit (IC) captured using an x-ray to demonstrate that IC authentication can be achieved when few photons are available in the scene. In the third part of this dissertation, three-dimensional (3D) imaging and object recognition in low light environments using real experimental data will be discussed. Afterwards, an object recognition framework for 3D reconstructed images obtained from elemental images taken in low light conditions using Convolutional Neural Networks will be introduced.