Target Identification via Laser Material Interaction
Nanohmics is developing an approach to remotely identify the composition of objects based on the characteristics of scattered light. A custom instrument (scatterometer) to measure the polarization dependent, Bidirectional Scattering Distribution Function (BSDF) has been developed in conjunction with a theoretical model that links the scattering function to the surface topology. In addition to a growing BSDF measurement services business, Nanohmics is pursuing the application of light scatter measurements in the areas of detection of fatigue in structural or machinery parts, non-contact detection of very small vibrations or motions, determination of process completion in semiconductor wafer processing, and monitoring of chemical processes in production plants.
The custom scatterometer is currently being utilized to support the development of enhanced aim point capabilities for the Air Force’s Advanced Tactical Laser (ATL) - a platform for high-precision strike missions utilizing a laser weapon mounted on a tactical aircraft. In normal operation, boresighted cameras provide the system with high-resolution imagery for the identification and discrimination of targets of interest. The ATL will allow precision, near-instantaneous delivery of laser energy to these targets from a long stand-off distance, with enough power density to melt metal. The ATL will thus allow the disability of vehicles, command and communications infrastructure, or even missiles, without substantial collateral damage or casualties. In situations where there is some uncertainty about the visual identification of targets, the coherent nature of the radiation from the ATL allows a more in-depth determination of what is being targeted by probing the material and surface characteristics of the irradiated object. By observing the polarization aspects of the back scattered light from an illuminated object, one can determine the approximate engagement geometry and material. The total amount of light detected as a function of engagement angle then allows classification of the object based on its surface roughness and texture. The dynamic aspects of both the polarization and detected scatter allows detection of object motion or rotation, providing further opportunity for classification of the object’s material, surface finish, and threat probability.
Clearly, it is desirable to gain a quantitative understanding of the laser scattering characteristics when various targets are illuminated. This information can be used as the basis for the development of accurate target identification and discrimination algorithms. To accomplish this, Nanohmics is developing an approach to the laser scattering problem from a theoretical and experimental perspective simultaneously. The fundamental physics of the vector scattering problem are being utilized to develop insight and expectations regarding the scatter characteristics of a wide variety of potential target materials, shapes, and surface characteristics and at a number of different wavelengths. In addition, scatterometer measurements are being made to verify the predictions of the analytical model and to demonstrate the performance of candidate target discrimination algorithms.
The nature of the laser-target interaction for realistic targets of interest is best described using a vector scattering theory that is valid for “rough” objects – here rough is a relative term, and denotes that the topographical features of the object’s surface are of the order of, or larger than, the illuminating wavelength. The power spectrum of surface topology and gross shape will determine the coarse features of the object’s bidirectional reflectance distribution function (BRDF), and the composition (and resulting optical constants n, k) will determine how the incident polarization is modified. Speckle information, although perhaps difficult to obtain because of the dynamic nature of the illumination due to the intervening atmosphere between the beam director and object, will be a function of the object’s surface roughness and short-range order. All of these scatter characteristics are difficult to compute from first principles for the full 4-dimensional parameter space of general bi-directional scattering. Fortunately, there is some simplification due to the constraint that the scatter from potential targets will be viewed by sensors close to the source, i.e. the backscatter direction. Other theoretical considerations as well as the engagement-viewing symmetry, allow simplification of the identification problem to one that can be managed by a statistical vector classifier.