Microbolometer sensor model for performance predictions and real-time image generation of infrared scenes and targets

Robert A. Richwine; Yash R. Puri; Ashok K. Sood; Raymond S. Balcerak; Stuart Horn; Glenn Baker; Michael Wilson

doi:10.1117/12.801337

Translator Disclaimer

26 August 2008 Microbolometer sensor model for performance predictions and real-time image generation of infrared scenes and targets

Robert A. Richwine, Yash R. Puri, Ashok K. Sood, Raymond S. Balcerak, Stuart Horn, Glenn Baker, Michael Wilson

Author Affiliations +

Proceedings Volume 7055, Infrared Systems and Photoelectronic Technology III; 70550T (2008) https://doi.org/10.1117/12.801337
Event: Photonic Devices + Applications, 2008, San Diego, California, United States

Abstract

Presented is a comprehensive, physics-based model for microbolometer detector and sensor performance prediction. The model combines equations found in the literature and various standard models that generate NETD, MRTD, 3-D noise statistics and atmosphere characteristics (MODTRAN-based), with a comprehensive microbolometer model and HgCdTe model developed by the author to provide an end-to-end detector/FPA/sensor analysis and design tool, as well as a realistic image sequence generation tool. The model characterizes the individual pixel element based on the structure used, the various layer thicknesses, the electrical and thermal characteristics of the bolometer material and the biasing and readout circuit, and uses these results to calculate response and noise, NEP and NETD. The NETD, MTF and MRTD are predicted from the optics, detector and readout. Predicted NETD has been compared and verified with values found in literature, results from other models, and to uncooled camera measurements. The MRTD prediction has been verified with camera measurements and with standard industry MRTD model outputs. The model also calculates atmospheric path radiance and transmittance for horizontal paths based on MODTRAN outputs for the LWIR band at altitudes from 0 to 10km and ranges from 1 to 50km for assessments of air-to-air engagement SNR's. The model in matlab utililizes a 3-D noise model to provide accurate realistic imagery used to present realistic sensor images and to further validate the NETD and MRTD routines.(1) Images at 30Hz and 60Hz have been generated for visual assessment by the user and have mirrored industry model results and real-time camera images for MRTD's for the temporal noise case. The model's 3-D noise generation feature allows the prediction of MRTD vs. frequency under any 3-D noise combination. This model provides an end-to-end performance prediction tool useful in bolometer element design, readout design and for system level trade studies.

Citation Download Citation

Robert A. Richwine, Yash R. Puri, Ashok K. Sood, Raymond S. Balcerak, Stuart Horn, Glenn Baker, and Michael Wilson "Microbolometer sensor model for performance predictions and real-time image generation of infrared scenes and targets", Proc. SPIE 7055, Infrared Systems and Photoelectronic Technology III, 70550T (26 August 2008); https://doi.org/10.1117/12.801337

ACCESS THE FULL ARTICLE