09/18/2023
By Dalila Megherbi

John DiZoglio will defend his Ph.D. dissertation thesis in Computer Engineering, titled "FTIR Spectral-Response-Based Signal Processing and Data Analysis Methods for Enhancement of Microsphere-Lense-Augmented MWIR Type II SLS Photo Detectors Response” on Oct. 3, 2023, at 1:30 p.m.

Location: This will be a virtual defense via Zoom. Those interested in attending should contact Committee Chair at Dalila_Megherbi@uml.edu and John_Dizoglio@student.uml.edu at least 24 hours before the defense to request access to the meeting.

Committee Members:

  • Committee Chair/Advisor D. B. Megherbi, ECE Department, (CMINDS), UML
  • Xuejun Lu, ECE Department, UML
  • Kanti Prasad, ECE Department, UML
  • Imran Vakil, Air Force Research Laboratory, Wright Patterson Base

Abstract:
Infrared imaging systems are an important part of many different fields of study. In the Mid-Wave Infrared (MWIR) band one of the better performing systems is the type-II Strained Layer Superlattice detectors (SLS). The importance of such systems stems from the fact that they operate well through fog, smoke and dust, which makes them desirable in many applications including military and airborne surveillance, manufacturing, security, night vision, threat detection, and inspection to name a few.

Engineers of new hardware sensors face the challenge of keeping costs low and performance high. Software estimation and simulation of these systems allow engineers to make knowledgeable decisions before entering the hardware phase of their design. When developing infrared sensors, controllable aspects include the array size, detector pitch, detector diameter, material composition, and thickness. Software tools are used to simulate and analyze the sensors, providing valuable information to the designers and saving development costs with post-processing software. Using software during development allows a faster turnaround time, with better accuracy and results than redeveloping new hardware each iteration. These methods can be used to both characterize previously manufactured sensors, as well as to aid in the design of future devices. These processes work on a broad array of sensor types, from visible light, to infrared, though this research focuses on one type of infrared sensor.

This dissertation focuses specifically on digital signal processing techniques and algorithms on microsphere-lens enhanced MWIR SLS detectors. These sensors are typically tested using a Fourier Transform Infrared (FTIR) spectrometer which is a machine capable of determining the percent transmittance and transmission of a light generated from a blackbody source through a material. Type II SLS detectors operate at ~80˚ Kelvin and thus require external cooling. They can technically operate at higher temperatures, but this reduces the sensor's sensitivity, thus increasing the Noise-to-Signal Ratio (NSR). Additionally, the microsphere lenses specific material type, like polystyrene, can introduce undesirable spectral absorption and aberration in the microsphere enhanced SLS photodetector.

Our research focuses on using the proposed methods, algorithms and enhancements to improve the sensor’s sensitivity, as well as to detect, characterize and remove both noise and material spectral absorption. Key features of the proposed methods include an improvement of type II SLS photodetector’s area sensitivity and analysis of the NSR. Additionally, we automatically correct for microsphere-lens-enhanced detector spectral response information loss due to material absorption and microsphere misalignment. Furthermore, we analyzed the effect of the microsphere size, its material composition, and the adhesive used in the application of the sphere on the spectral response of the microsphere-enhanced detector.

The proposed methods include signal analysis and smoothing techniques such as Butterworth filtering, custom averaging filtering, and poly-fit analysis. Each of these algorithms are studied with varying degrees of aggressiveness so that an ideal balance between signal smoothing and loss of detector spectral information is found. A performance metric denoted by “area performance” is used to compare the efficacy of these algorithms. These techniques are used to develop a way to separately compare microsphere-lens-enhanced signal loss due to noise and material absorption. In addition to the signal processing and analysis techniques, we also show the effects of microsphere misalignment and the glue’s various refractive indices on the microsphere-enhanced detector using real experimental and simulated FTIR responses. The experimental results confirm the utility and value of the proposed methods and pave the way for exploration and development of new sensors in the future.

All interested students and faculty members are invited to attend the defense.