Object detection, a critical task in computer vision, has been revolutionized by Deep Learning technologies, especially convolutional neural networks (CNN). These techniques are increasingly deployed in infrared imaging systems for long-range target detection, localization, and identification. Its performance is highly dependent on the training procedure, network architecture and computing resources. In contrast, human-in-the-loop task performance can be reliably predicted using well-established models. Here we model the performance of a CNN developed for MWIR and LWIR sensors and compare against human perception models. We focus on tower detection relevant to vision-based geolocation tasks which present novel high-aspect ratio, unresolved and low-clutter scenarios.
KEYWORDS: Sensors, Cameras, Global Positioning System, Long wavelength infrared, Panoramic photography, Short wave infrared radiation, Mid-IR, Near infrared, MATLAB, Algorithm development
We built a multispectral data collection system and vehicle testbed for experimentation on vision-based geolocation. The data collection system includes a gimballed mount with VIS, NIR, SWIR, MWIR and LWIR sensors allowing us to compare simultaneous imagery of features or targets across all the atmospheric bands. For geolocation experiments, the testbed is equipped with a dual-GPS inertial measurement unit for true location and orientation, a CAN bus interface to pull vehicle speedometer and odometer data. It is also outfitted with a dual-GPU, rugged, edge computer used to control the system and collect data; the computer is pre-loaded with geospatial data (maps, tower positions, elevation data, etc.) necessary for tracking targets of interest or performing real-time geolocation estimates. 60% of the rear seat was replaced with an electronics rack which also houses a 3kW inverter providing power to all of the equipment. A weather-proof cable pass through was installed in the roof of the truck while a weather-proof enclosure provides wind and rain protection to the roof-mounted equipment. We present multispectral panoramic imagery of flat environments where cellular towers provide ideal references for geolocation and mountainous environments where the landscape and horizon topography provide viable geo-references. We will present an overview of the data collection modes, calibration procedures, and the driving data sets collected to date.
The ability to reliably and accurately ascertain a vehicle’s position is imperative for military operations as well as civilian and commercial navigation systems. Due to the susceptibility of GPS signals to RF spoofing and jamming, alternative means of vehicle self-localization are garnering substantial interest. Vision-based methods are among the most promising in environments with sufficiently distinguishable features such as towers, high-rise structures, and prominent identifiable topographical features. Here, we present a localization approach exploiting multiple spectral bands to identify key prominent scene features and determine vehicle position relative to those features to calculate a global vehicle position and heading. We employ geometric dead-reckoning using visible and LWIR imagery to quantify positional accuracy that is achievable with these bands. We utilize image recognition algorithms to identify features and map these into useful parameters for position extraction, leveraging geospatial data when possible.
In this work, we report using an optical tweezers system to study the light-matter interaction and gradient optical forces of porous silicon nanoparticles. The particles are fabricated by first electrochemically etching a multi-layer porous film into a silicon wafer and then breaking up the film through ultrasonic fracturing. The particles have average pore diameters ranging from 20-30 nm. The fabricated batches of particles have diameters between approximately 100- 600nm. After fabrication, the particles are size-sorted by centrifugation. A commercially available optical tweezers system is used to systematically study the optical interaction with these nanoparticles. This work opens new strategic approaches to enhance optical forces and optical sensitivity to mechanical motion that can be the basis for future biophotonics applications.
Porous materials offer several advantages for chemical and biomolecular sensing applications. In particular, nanoscale
porous materials possess a very large reactive surface area to facilitate the capture of small molecules, and they have the
capability to selectively filter out contaminant molecules by size. This paper will provide an overview of the fabrication,
functionalization, and application of porous silicon thin films and waveguides, as well as porous gold templates, for the
detection of small chemical and biological molecules. Issues of efficient molecule infiltration and capture inside porous
materials, binding kinetics in nanoscale pores, the influence of pore size on small molecule detection sensitivity, and the
new nanoscale patterning technique of Direct Imprinting of Porous Substrates (DIPS) will be addressed. Additionally, a
novel application of porous silicon for detection of x-ray radiation will be introduced.
We will discuss the epitaxial growth, characterization, and application of a new set of ternary cubic oxide
semiconductor compounds, ZnxMg1-xO and NiyMg1-yO, offering a new route towards deep-UV optical devices.
Results demonstrating bandgap tunability and excellent thin film quality will be presented validating the potential of
these alloys in the 200 - 350 nm region. Significantly, we have successfully fabricated MSM solar blind detectors
using both ternary alloys, demonstrating operation in the solar blind region without external opal filters.
Oxide based compounds have been of increasing interest for wide bandgap, deep ultraviolet optoelectronics. While high
Al content AlGaN has enabled many UV-DUV technologies, it suffers inherent drawbacks including difficulty achieving
increasing Al incorporation, high threading dislocation densities and challenges in bandgap engineering due to
polarization and piezoelectric effects. Here we present two wide bandgap cubic oxide compounds, ZnMgO and NiMgO,
that offer advantages over AlGaN for deep ultraviolet (DUV) applications. NixMg1-xO and ZnxMg1-xO are both direct
band gap, cubic rocksalt (B1) semiconductors with bandgaps in the UV-DUV spectral regions, offering alternatives
without the aforementioned drawbacks associated with AlGaN. Here we present NixMg1-xO and ZnxMg1-xO thin films
grown by plasma-assisted MBE on lattice matched MgO substrates as a novel means by which to realize DUV detection
devices. In both systems we have shown the films to exhibit abrupt, continuously tunable absorptions edges over their
respective bandgap ranges. NixMg1-xO films were varied compositionally from x=0 to 1, realizing bandgaps from 3.5 to
7.8 eV. ZnxMg1-xO films were similarly varied over the entire B1 range of the ternary (0<x<0.42) and show bandgap
tunability from ~5 to 7.8 eV. All films are characterized through Rutherford backscattering (RBS), x-ray diffraction
(XRD), atomic force microscopy (AFM) measurements and optical transmission. Significantly, we have successfully
fabricated solar blind detectors in both categories and highlight the results from NixMg1-xO here.
GaAs-based PIN detectors with mesa sizes 1, 2.5, 5, 7.5 and 10 mm were fabricated and characterized for alpha particle
response using a Po-210 alpha source. By decoupling the neutron conversion process of a proximity moderator, we were
able to directly probe the alpha response characteristics of the PIN detectors as a function of device area. Dark current
levels in the PIN detectors ranged from 6.1 to 9.5 pA at zero bias. The dark current values were higher for larger devices
and a linear relationship between mesa size and dark current was observed. The PIN detectors were found to have a
strong alpha response of up to 5 nA/mm2 with a linear relation between the response current and mesa area. The
measured responsivity of the detectors was 0.014 A/W. The average device efficiency was determined to be 31.5%.
Using the measured alpha response properties of the GaAs PIN diodes one is able to select the optimal device area for a
given moderator and application specific neutron flux.
Femtosecond ablation has several distinct advantages: the threshold energy fluence for the onset of damage and ablation is orders of magnitude less than for traditional nanosecond laser machining, and by virtue of the rapid material removal of approximately an optical penetration depth per pulse, femtosecond machined cuts can be cleaner and more precise than those made with traditional nanosecond or longer pulse lasers. However, in many materials of interest, especially metals, this limits ablation rates to 10-100 nm/pulse. We present the results of using multiple pulse bursts to significantly increase the per-burst ablation rate compared to a single pulse with the same integrated energy, while keeping the peak intensity of each individual pulse below the air ionization limit. Femtosecond ablation with pulses centered at 800-nm having integrated energy of up to 30 mJ per pulse incident upon thin gold films was measured via resonance frequency shifts in a gold-electrode-coated quartz-crystal oscillator. Measurements were performed using Michelson-interferometer-based burst generators, with up to 2 ns pulse separations, as well as pulse shaping by programmable acousto-optic dispersive filter (Dazzler from FastLite) with up to 2 ps pulse separations.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.