While some instrument requirements are levied on a per-pixel basis, efficiencies and economies can be gained by testing them in parallel. Furthermore, the use of detector arrays as imagers with extended targets enables the derivation of geometrical information from select pixels in each image, and its propagation to neighboring pixels. We discuss the implementation of one such test regime for the Operational Landsat Imager (OLI) at Ball Aerospace and Technologies Corp. This enabled rapid measurement of spatial parameters, including Edge Response Function and aliasing, for all of the nearly 70,000 active pixels of the focal plane assembly with reduced reliance on the precision and stability of the supporting equipment. The derived geometrical information enabled us to replace a step-stare testing of individual pixels with a continuous scan of the entire assembly, without demanding precision motion or introducing noise from variations in the scan velocity. Three complete scans were performed in under 30 hours.
3D imaging LADARs have emerged as the key technology for producing high-resolution imagery of targets in 3-dimensions (X and Y spatial, and Z in the range/depth dimension). Ball Aerospace & Technologies Corp. continues to make significant investments in this technology to enable critical NASA, Department of Defense, and national security missions. As a consequence of rapid technology developments, two issues have emerged that need resolution. First, the terminology used to rate LADAR performance (e.g., range resolution) is inconsistently defined, is improperly used, and thus has become misleading. Second, the terminology does not include a metric of the system’s ability to resolve the 3D depth features of targets. These two issues create confusion when translating customer requirements into hardware. This paper presents a candidate framework for addressing these issues. To address the consistency issue, the framework utilizes only those terminologies proposed and tested by leading LADAR research and standards institutions. We also provide suggestions for strengthening these definitions by linking them to the well-known Rayleigh criterion extended into the range dimension. To address the inadequate 3D image quality metrics, the framework introduces the concept of a Range/Depth Modulation Transfer Function (RMTF). The RMTF measures the impact of the spatial frequencies of a 3D target on its measured modulation in range/depth. It is determined using a new, Range-Based, Slanted Knife-Edge test. We present simulated results for two LADAR pulse detection techniques and compare them to a baseline centroid technique. Consistency in terminology plus a 3D image quality metric enable improved system standardization.
A new technology for performing high-precision Stokes polarimetry is presented. One traditional Stokes polarimetry configuration relies on mechanical devices such as rapidly rotating waveplates that are undesirable in vibration-sensitive optics experiments. Another traditional technique requires division of a light signal into four components that are measured individually; this technique is limited to applications in which signal levels are sufficient that intensity reduction does not diminish the signal-to-noise ratio. A new technology presented here is similar to the rotating waveplate approach, but two liquid crystal variable retarders (LCVR’s) are used instead of waveplates. A Stokes polarimeter instrument based on this technology has been made commercially-available. The theory of operation is detailed, and an accuracy assessment was conducted. Measurement reproducibility was verified and used to produce empirical estimates of uncertainty in measured components of a Stokes vector. Uncertainty propagation was applied to polarization parameters calculated from Stokes vector components to further the accuracy assessment. A calibrated polarimeter measures four Stokes components with 10-3 precision and average predicted uncertainties less than ±2x10-3. An experiment was conducted in which the linear polarization angles were measured with a LC polarimeter and with a photodiode for comparison. Observed discrepancies between polarization angle measurements made with a polarimeter and those made with a photodetector were nominally within ±0.3°.