|M.Sc Student||Tomer Merhav|
|Subject||An Electro-Optic Flash Detection System in the Visible|
|Department||Department of Electrical Engineering||Supervisor||Professor Emeritus Nemirovsky Yael|
|Full Thesis text|
Flash detection is the process of extracting information about a short-lived physical event from light emitted by it. Various fields of science and technology use flash detection as a means for analyzing certain physical phenomena. Examples range from combat sniper detection systems where the rifle muzzle flash is of interest, to Fluorescence Correlation Spectroscopy (FCS) where photons emitted by a burst-like event, correspond to the transfer of a single molecule. These two very different examples share common properties, which dictate demanding requirements of the detection system, namely high sensitivity accompanied by high sampling frequency. However, sniper detection poses additional challenges because it is performed in an uncontrolled environment. Electro-optic detection systems achieve the highest signal-to-noise ratios (SNR) when working in the Mid Wave Infrared waveband, but this constraint dictates a high price tag together with other limitations such as weight and mobility. Sniper detection in the visible waveband is considered impractical due to a low SNR, and thus has not been thoroughly studied. Detection systems in the visible waveband are required to incorporate high-frequency high-sensitivity imagers, in-order to enable flash detection in an outdoor environment. This research concentrates on utilizing recent advances in visible waveband imaging technologies in order to characterize a detection solution, which is both low-power and low-price. Such a solution will be based on mature CMOS technology, and will be adequate for any flash detection task, provided the correct operational parameters. For muzzle flash detection, Avalanche photodiodes are employed, inorder to amplify the faint visible wavelength signal of the muzzle flash. Experiment show promising results, both in detecting the muzzle flash and rejecting false alarms due to background clutter.